Datasets:

WINGNUS
/

ACL-OCL

	{
	"paper_id": "J76-3001",
	"header": {
	"generated_with": "S2ORC 1.0.0",
	"date_generated": "2023-01-19T02:52:18.216083Z"
	},
	"title": "A C M I EMPLOYMENT REGISTER AT COMPUTER SCIENCE CONFERENCE",
	"authors": [
	{
	"first": "A",
	"middle": [
	"Hood"
	],
	"last": "Roberts",
	"suffix": "",
	"affiliation": {},
	"email": ""
	},
	{
	"first": "James",
	"middle": [
	"S"
	],
	"last": "Megginson",
	"suffix": "",
	"affiliation": {},
	"email": ""
	},
	{
	"first": "I",
	"middle": [
	"I"
	],
	"last": "Qubine",
	"suffix": "",
	"affiliation": {},
	"email": ""
	},
	{
	"first": "Raymond",
	"middle": [
	"D"
	],
	"last": "Gumb",
	"suffix": "",
	"affiliation": {},
	"email": ""
	},
	{
	"first": "Friday",
	"middle": [],
	"last": "October",
	"suffix": "",
	"affiliation": {},
	"email": ""
	},
	{
	"first": "",
	"middle": [],
	"last": "Registration",
	"suffix": "",
	"affiliation": {},
	"email": ""
	},
	{
	"first": "George",
	"middle": [
	"E"
	],
	"last": "Heidorn",
	"suffix": "",
	"affiliation": {},
	"email": ""
	},
	{
	"first": "B",
	"middle": [
	"A"
	],
	"last": "Sheil",
	"suffix": "",
	"affiliation": {},
	"email": ""
	},
	{
	"first": "Jerry",
	"middle": [
	"R"
	],
	"last": "Hobbs",
	"suffix": "",
	"affiliation": {},
	"email": ""
	},
	{
	"first": "Harry",
	"middle": [
	"C"
	],
	"last": "Bunt",
	"suffix": "",
	"affiliation": {},
	"email": ""
	},
	{
	"first": "John",
	"middle": [],
	"last": "Bennett",
	"suffix": "",
	"affiliation": {},
	"email": ""
	},
	{
	"first": "-Madison",
	"middle": [],
	"last": "Frame",
	"suffix": "",
	"affiliation": {},
	"email": ""
	},
	{
	"first": "Joyce",
	"middle": [
	"B"
	],
	"last": "Friedman",
	"suffix": "",
	"affiliation": {},
	"email": ""
	},
	{
	"first": "F",
	"middle": [],
	"last": "Michigan",
	"suffix": "",
	"affiliation": {},
	"email": ""
	},
	{
	"first": "Ljilliam",
	"middle": [
	"H"
	],
	"last": "Paxton",
	"suffix": "",
	"affiliation": {},
	"email": ""
	},
	{
	"first": "Stanford",
	"middle": [],
	"last": "Research",
	"suffix": "",
	"affiliation": {},
	"email": ""
	},
	{
	"first": "Inst",
	"middle": [],
	"last": "Stanley P E T R I C K",
	"suffix": "",
	"affiliation": {},
	"email": ""
	},
	{
	"first": "Ibm",
	"middle": [],
	"last": "Research",
	"suffix": "",
	"affiliation": {},
	"email": ""
	},
	{
	"first": "Eric",
	"middle": [],
	"last": "Van Utteren",
	"suffix": "",
	"affiliation": {},
	"email": ""
	},
	{
	"first": "",
	"middle": [],
	"last": "Philips",
	"suffix": "",
	"affiliation": {},
	"email": ""
	},
	{
	"first": "William",
	"middle": [
	"C"
	],
	"last": "Mann",
	"suffix": "",
	"affiliation": {},
	"email": ""
	},
	{
	"first": "James",
	"middle": [
	"A"
	],
	"last": "Levin",
	"suffix": "",
	"affiliation": {},
	"email": ""
	},
	{
	"first": "James",
	"middle": [
	"A"
	],
	"last": "Moore",
	"suffix": "",
	"affiliation": {},
	"email": ""
	},
	{
	"first": "Allen",
	"middle": [],
	"last": "Munro",
	"suffix": "",
	"affiliation": {},
	"email": ""
	},
	{
	"first": "Charles",
	"middle": [
	"J"
	],
	"last": "Rieger",
	"suffix": "",
	"affiliation": {},
	"email": ""
	}
	],
	"year": "",
	"venue": null,
	"identifiers": {},
	"abstract": "Bath the Current Bibliography and t h e promised index t o AJCL are absept from the present packet, because of the difficulties inherent in o p e r e t i~n s based on voluntary effort. American, Journal.of Computational Linguistics is published by the Center for Applied Linguistics for the Assaciation for Computational Linguistics.",
	"pdf_parse": {
	"paper_id": "J76-3001",
	"_pdf_hash": "",
	"abstract": [
	{
	"text": "Bath the Current Bibliography and t h e promised index t o AJCL are absept from the present packet, because of the difficulties inherent in o p e r e t i~n s based on voluntary effort. American, Journal.of Computational Linguistics is published by the Center for Applied Linguistics for the Assaciation for Computational Linguistics.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Abstract",
	"sec_num": null
	}
	],
	"body_text": [
	{
	"text": "The idea of attaching a modifier t o its nearest potential ",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "AN E A S I L Y COMPUTED M E T R I C FOR R A N K I N G u A L T E R N A T I V E PARSES GEORGE E. HEIDORN I B M RESEARCH",
	"sec_num": null
	},
	{
	"text": "branching examples, i t can a p p l y e q u a l l y w e l l t o l e f t -b r a n c h i n g structures.)",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "r a t e s i t w i t h r i g h t -",
	"sec_num": null
	},
	{
	"text": "I n t h e s e n t e n c e , \"Are t h o s e i n v o i c e s produced from t h e o r d e r s processed by t h e system i n New York?\", t h e p r e p o s it i o n a l phxase \" i n New York\" could p o t e h t i a l l y modify \"system\", ''processed1', \"orders\" , \"produced\" and \" i n v o i c e s \" , b o t h s y n t a ct i c a l l y and s e m a n t i c a l l y . S i m i l a r l y , \"by the system\" has f o u r possibilities and \"processed\" has two. According t o the h e u r i s t i c s t a t e d above, t h e p r e f e r r e d a n a l y s i s i s : \" i n New York\" m o d i f i e s \"system\", \"by t h e system. . . I ' m o d i f i e s \"processed\", and \"processed. . . as the lexemes of a n a t u r a l language). \u00b6?he basic idea i s t o m o d e y frequency ordered binary search (FOBS) trees t o cont a i n two key values--a node value which i d e n t i f i e s t h e key which resides a t t h a t node (as i n a conventional binary search tree), and a s p l i t value which gives the l a r g e s t key value t o be found i n the l e f t subtree. ( I f the keys a r e expensive t o s t o r e in t h e t r e e nodes, or c o s t l y t o compare, they may be hashed i n t o integers without a f f e c t i n g t h e algorithm. )",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "r a t e s i t w i t h r i g h t -",
	"sec_num": null
	},
	{
	"text": "Searching an MS t r e e proceeds as i n a binary t r e e except that the decision t o go l e f t o r r i g h t from a node whose node v a l u e does not match the current key i s made by comparing t h e current key t o the s p l i t , r a t h e r than t o the node, value. The use of t w o different values allows one t o prevent t h e search tree from being unbalanced as FOBS t r e e s become when a high frequency key has an.extreme key value. In f a c t , by s e l e c t i n g the median of the key values of a node's descendants a s i t s split value, one can force the search t r e e t o be p e r f e c t l y balancedo-which both allows a highly space e f f i c i e n t representation of the t r e e , and achieves high speed search.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "r a t e s i t w i t h r i g h t -",
	"sec_num": null
	},
	{
	"text": "The average cost per search in a n MS t r e e , like a FOBS tree, depend's on b o t h t h e frequency distribution of the keys, and t h e o r d e r i n g relabion on them. Performance analyses f o r specEic f r e q u e n c y distributions and key orderings of i n t e r e s t are presented. and it is shown that, unlike FOBS trees, PIS t r e e search t i m e i s l o g n bounded and is very stable around a value that can be shown t o be o p t i m a l f o r any binary tree search p r o c e d u r e . Furthermore, i n addition t o being substant i a l l y f a s t e r than \"optimum\" b i n a r y t r e e search, an EIS tree c a n be b u i l d for a given ordering r e l a t i o n and s e t of frequenc i e s in time n log n, as opposed t o n 2 A discussion of the a p p l i c a t i o n of MS trees to dictionary lookup for English is p r e s e n t e d , and the perfbrmance obtained.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "14th ACL Meeting",
	"sec_num": null
	},
	{
	"text": "is contrasted with hash and trie solutions.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "14th ACL Meeting",
	"sec_num": null
	},
	{
	"text": "JERRY R. HOBBS",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "TRANSLATING 'WELL-WRITTEN' ALGORITHM DESCRIPTIONS INTO CODE",
	"sec_num": null
	},
	{
	"text": "In this paper we describe a system for the semantic analysis of \"well-written\" algorithm descriptions in English, such as one finds in Knuth ( 3 ) , and t h e translation of t h e resulting structures into a program in a higher-level programming language, such as P L f I . There are two approaches one might take toward the sublanguage o f algorithm descript i o n s . One may t r e a t t.he E n g l i s h as a d e v i a n t p r o g r a m i n g language and try to smooth out t h e d e v i a t i o n s . Or one may v i e w t h e English as a well-behaved, h i g h l y regular form of natural language, and a p p l y to the t e x t the sophisticated t e c h n i q u e s of analysis that have been developed for less t r a c t a b l e sorts of t e x t s . W e have taken the l a t t e r approach, Semantic analysis is done by a system t h a t has been d eveloped for and is being tested on a variety of natural language texts (1). It takes as i n p u t the t e x t i n a p r e d i c a t e calculus-like n o t a t i o n produced by a syntactic front end (4, 2). I t a p p l i e s s e m a n t i c operations which access a data base of \"world knowledge\" inferences that are drawn selectively, 14th ACL Meeting I t i s intended t h a t these t r a n s l a t e not i n t o successive i ns t r u c t i o n s , but i n t o a case statement. This i s keyed by t h e",
	"cite_spans": [
	{
	"start": 135,
	"end": 146,
	"text": "Knuth ( 3 )",
	"ref_id": null
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "CI TY COLLEGE O F CUNY",
	"sec_num": null
	},
	{
	"text": "Contrast p a t t e r n . (The top-lev-el predicate i s \"imply\". )",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "CI TY COLLEGE O F CUNY",
	"sec_num": null
	},
	{
	"text": "S i m i l a r l y , consider the sentences (1) \"Decrease N by 1,. and if i t i s zero, r e s e t i t t o MAX.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "CI TY COLLEGE O F CUNY",
	"sec_num": null
	},
	{
	"text": "(2) \"Decrease N by 1, but i f i t i s zero, r e s e t i t t o MAX.\" I n (1) the t e s t f o r zero comes a f t e r decreasing N , in ( 2 ) before. This i s because i n ( 2 ) , \"but\" invokes the Contrast p a t t e r n , f o r c i l g us t o recover the implicit \" I f N is not zero\"",
	"cite_spans": [
	{
	"start": 126,
	"end": 138,
	"text": "N , in ( 2 )",
	"ref_id": null
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "t I",
	"sec_num": null
	},
	{
	"text": "before \"Decrease\".",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "t I",
	"sec_num": null
	},
	{
	"text": "( 2 ) is t r a n s l a t e 3 i n t o a c a s e statement.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "t I",
	"sec_num": null
	},
	{
	"text": "This frequently r e s o l v e s anaphora a s a byproduct.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Knitting: Redundancies are recognized and merged",
	"sec_num": "3."
	},
	{
	"text": "Antecedents not Found by k n i t t i n g are searched for by t r y i n g t o maximize t h e redundancy of the t e x t . I n \"Decrease N by J , and i f it i s zero, r e s e t i t t o MAX.\" ltN\" is chosen over \"J\" a s the antecedept of \"it1' s i n c e i t i s N whose value i s subject-t o change. i n s t a n c e , a t r e e i s , i n p a r t , a s e t of nodes, a node i s a s e t of f i e l d s , and a f i e l d may be a number. \"Set1' and \"number\"",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": ". Resolving anaphora :",
	"sec_num": "4"
	},
	{
	"text": "are p r i m i t i v e s .",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": ". Resolving anaphora :",
	"sec_num": "4"
	},
	{
	"text": "To decrease N by 1 is t o cause N t o be equal t o one l e s s than before, and \"cause t o be equal to\" is a p r i m i t i v e .",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": ". Resolving anaphora :",
	"sec_num": "4"
	},
	{
	"text": "What i s produced i s then given t o t h e t a s k component which per forms two Punc t ions :",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Semantic Analysis augments and interrelates the texk",
	"sec_num": null
	},
	{
	"text": "1. The temporal succession r e l a t i o n s discovered by o p e r a t i o n 2 may form onrly a partial ordering on actions. A ltnear ordering is imposed.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Semantic Analysis augments and interrelates the texk",
	"sec_num": null
	},
	{
	"text": "t r a n s l a t e the semantic representatidn i n t o code. A sebt translates i n t o an array declaration, a s e t of s e t s i n t o a two-dimensional a r r a y . \"Carse t o be equal to\" t r a n s l a t e s i n t o an assignment s t a t e m e n t . An automatic programming system that converses with a u s e r in a natural language must be able to generate explanations of portions of the program being produced. In the system that wc are developing for producing customized business application programs, these explanations.must be given in application terms rather than program terms. For example. the system might sey , \"The quantity shipped is the lesser of the quantity o r d e r e d and the quantity available,\" rather than \"QS=MIN(QO,QA)\".",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": ". Sobstitution rules a r e applied t o primitives t o",
	"sec_num": "2"
	},
	{
	"text": "One could certainly imagine generating such sentences by having an inventory of sentence patterns and name phrases and using a fill-in-the-blanks approach. However, in addition to not being very interesting, such an approach would certainly not provide the same flexibility of expression that a more linguistically motivated approach would. Also, it seems reasonable to do this generation by utilizing information that,. the system already has for understanding the user's English utterances. This uncertainty is reflected in the appearance of a variety of vague terms, such as \"substance\", \"quantity\", \"matter\", \"bits\", \"pieces1', \"stuff\", etc. in recent literature on the \"object\"), features like \"emotion, \" \"authority, \" \"concept , 11",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Bibliography",
	"sec_num": null
	},
	{
	"text": "and \"avocation\" have been g e n e r a t e d . F u r t h e r work will look i n t o t h e use of f e a t u r e s i n o t h e r tasks of u n d e r s t a n d i n g , cializing and extending their existing ability to communicate.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Bibliography",
	"sec_num": null
	},
	{
	"text": "To design systems which interact effectively with humans, we need to better understand how people communicate with each other.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Bibliography",
	"sec_num": null
	},
	{
	"text": "In particular, we need to view human communication as a problem-solving activity, in which the people so engaged are using language as a method to achieve certain of their goals.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Bibliography",
	"sec_num": null
	},
	{
	"text": "The existing research into natural language has not focused on this aspect, so from our point of view the work so far has been fragmentary and hard to use, particularly in terms of h e l p i n g to enhance man-machine corn-nication.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Bibliography",
	"sec_num": null
	},
	{
	"text": "Constructing a useful model of human communication is an extremely complex task. We view t h e controlling of this complexity (without sacrificing utility) as the central problem in designing our research approach. In response to this challenge, our research methodology contains some innovations, specifically intended to maintain control on the complexity, while retaining the usefulness of our results for designers of man/machine communication systems. proposed higher level units, like \"scripts\" or \"frames\".",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Bibliography",
	"sec_num": null
	},
	{
	"text": "There a r e a l s o higher l e v e l u n i t s called Dialogue-games, specifications concerning t h e topic of discussion, t h e particip a n t s , the goals they a r e trying to achieve with a given type have found that these dialogue forms are identified i n conversation by utterances that attempt to establish these parameters.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Bibliography",
	"sec_num": null
	},
	{
	"text": "Once a Dialogue-game has been activated as possibly the cornmunication form being proposed for a dialogue, the Dialogue-game",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Bibliography",
	"sec_num": null
	},
	{
	"text": "Processor operates on i t to verify that the parameters are properly specified, and then to establish the subgoals that are specified in LTM as the components of the particular A few brief comparisons with other natural language understanding techniques will be included.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Bibliography",
	"sec_num": null
	},
	{
	"text": "14th ACL Meeting",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Bibliography",
	"sec_num": null
	},
	{
	"text": "A s y s t e m for representing and using knowledge of cause and effect, called Cormnonsense Algorithms (CSA), will be described. The CSA project, ongoing for about a year now, has CSA theory incorporates processes which are demon-like and processes which are more planful, and the CSA theory identifies how and why the planner and the population of demons interact. The presentation will cover the organization and access techniques for large numbers of CSA causal patterns in these two-modes (planful and demoiic) , drawing on scenarios which illustrate the level of sophistication of the current implementation. One scenario will be taken from the children's 2) Redundancies a r e recognized and merged ( k n i t t i n g ) .",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "UNIVERSITY OF MARYLAND",
	"sec_num": null
	},
	{
	"text": "as",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "UNIVERSITY OF MARYLAND",
	"sec_num": null
	},
	{
	"text": "3 ) Anaphora a r e resolved.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "UNIVERSITY OF MARYLAND",
	"sec_num": null
	},
	{
	"text": "The operations work by accessing a data base of world knowledge inferences, which a r e drawn s e l e c t i v e l y i n response t o the operations, and which carry a measure of their salience that varies with the context. As analysis proceeds, a tree-like structure is constructed for the paragraph, with subordinating relations building the tree downward and coordinating links building it to the right.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "UNIVERSITY OF MARYLAND",
	"sec_num": null
	},
	{
	"text": "A previous sentence is \"eligible\" if it is on the right frontier of this tree. For example, in a text S1 S2 Sg, if S2 is a disguised relative clause on S1, then Sg may be coordinated with either S1 or S2, but if Sg succeeds S1 temporally, then",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "UNIVERSITY OF MARYLAND",
	"sec_num": null
	},
	{
	"text": "Sg may Be coordinated only with S p .",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "UNIVERSITY OF MARYLAND",
	"sec_num": null
	},
	{
	"text": "The heart of the operation is an ordered heuristic search of the data base for desired inferences. These 14th ACL Meeting desired inferences are kept on a goal list, and have strengths associated with them. The s t r e n g t h of a goal depends on the type of text and the patterns found previously in the text.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "UNIVERSITY OF MARYLAND",
	"sec_num": null
	},
	{
	"text": "Also, the presence of conjunctions and some other elements advance certain patterns--'>.cndu promotes Temporal Succession and a repetition of the previo~sly recognized pattern, a dash and \"l.e.\" promote Paraphrase, and the articLe \"this\" frequently signals a disguised relative clause. The evaluation function which orders the heuristic search is based on the srength of the goal, the length of the chain of inference, and the current saliencerof inferences in the chain.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "UNIVERSITY OF MARYLAND",
	"sec_num": null
	},
	{
	"text": "When a partial match is found, the difference, or the remainder of the goal, is placed high on a goal list for subsequent processing. Consider the text \"Republicans. are discouraged about their prospects",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "UNIVERSITY OF MARYLAND",
	"sec_num": null
	},
	{
	"text": "The party chairman is convinced that many GOP congressaen will lose their bid for reelection. ? 1",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "UNIVERSITY OF MARYLAND",
	"sec_num": null
	},
	{
	"text": "Suppose for simplicity \"be discouraged\" decomposes into \"believe something bad will happen\"",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "UNIVERSITY OF MARYLAND",
	"sec_num": null
	},
	{
	"text": "We have a partial match with the Example pattern since the secdnd sentence asserts that a particular Republican believes something.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "UNIVERSITY OF MARYLAND",
	"sec_num": null
	},
	{
	"text": "Hence a principal goal for our processing of the \"that\" clause is to show that what is believed is than an event bad for a Republ-ican will occur. This is shown by accessing the fact about a political party that one of its purpdses is to win elections .",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "UNIVERSITY OF MARYLAND",
	"sec_num": null
	},
	{
	"text": "Goals generated by a p a r t i a l match a r e also passed on to subsequent sentences. I n a Newsweek paragraph analyzed,",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "14th ACL Meeting",
	"sec_num": null
	},
	{
	"text": "t h e f i r s t sentence S1 asserts a change. The next three sentences S2 S g S4 assert stages along the course of t h i s change. S2 is matched w i t h the i n i t i a l s t a t e of the change in S1, thus generating as a goal the matching of Sg and S4",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "14th ACL Meeting",
	"sec_num": null
	},
	{
	"text": "with succeeding s t a t e s . Finally, the structure \"s2 -then S3",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "14th ACL Meeting",
	"sec_num": null
	},
	{
	"text": "then S4\" is taken as an Example of S1.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "14th ACL Meeting",
	"sec_num": null
	},
	{
	"text": "This work has significance f o r future attempts t o summarize paragraphs automatically. For, i n s t a n c e , i n a t e x t , t i e d together by the Example pattern, t h e sentence \"exampled\" must be a major p a r t of the summary. Unfortunately , much of t h e money spent on machine translation projects was applied t o theoretical research r a t h e r than being used for translation--which was often disparaged as being In the ten years since the ALPAC report, there has been considerable development in computer technology and i n linguistics.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "14th ACL Meeting",
	"sec_num": null
	},
	{
	"text": "The state of the a r t has advanced i n both fields to the point where a new synthesis is now possible, which could produce greatly i mproved translations on a more cost-effective basis. (Unfortunately, The time has m w come for a new effort in MT to be undertaken. Properly conducted, such an effort would not only improve the quality gnd efficiency of t r a n s l a t i o n , but would add to our knowledge QE substantive u n i v e r s a l s and semantics, as well as",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "American Journal of Computational Linguistics",
	"sec_num": null
	},
	{
	"text": "THE FUTURE O F MT",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "American Journal of Computational Linguistics",
	"sec_num": null
	},
	{
	"text": "deepen our understanding of pa~ticular languages. MT can make an important contribution to the building of the information base on which the growth of linguistic theory must depend, a t the same time that it produces a result of great practical value. solved with the help of context a n a l y s i s .",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "American Journal of Computational Linguistics",
	"sec_num": null
	},
	{
	"text": "The fourth report at the plenary session was read by Dr. wozds inflecting uniformly. This method makes it possible to reduce a reverse dictionary of word-forms to the size of an ordinary d i c tione ry .",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "American Journal of Computational Linguistics",
	"sec_num": null
	},
	{
	"text": "In t h i s s e s s i o n the p a r t i c i p a n t s i n the Seminar discussed the principles of compilation of MT dictionaries and came t o t h e conclusion that it is necessary:",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "American Journal of Computational Linguistics",
	"sec_num": null
	},
	{
	"text": "to considel: an MT system as a whole in which all parts are interlocked, while the dictionary is its principal component;",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "American Journal of Computational Linguistics",
	"sec_num": null
	},
	{
	"text": "to design M systems for sublanguages and consequently cofnpile automatic dictionaries for well-defined topics;",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "American Journal of Computational Linguistics",
	"sec_num": null
	},
	{
	"text": "to pay special attention to strict selection of entries for automatic dictionaries, and for this purpose compile various auxiliary dictionari'es and concordances;",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "American Journal of Computational Linguistics",
	"sec_num": null
	},
	{
	"text": "to increase the volume of diverse syntactic, semantic, and lexical information in automatic dictionaries;",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "American Journal of Computational Linguistics",
	"sec_num": null
	},
	{
	"text": "to attach great importance to the structure of automatic dictionaries. According to a number of the participants of the seminar, the optimum structure is a combination of a dictionary of stems and a dictionary of wcrdforrps. The dictionary of wordforms contains more frequent words and the dictionary of stems contains the less frequent words. In case several trees correspond to the analyzed phrase then the The reports submitted to th@ section on semantic analysis touched on some narrow probleqs. ",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "American Journal of Computational Linguistics",
	"sec_num": null
	},
	{
	"text": ". Hobbs, J . R . and Grishman, Ralph. The automatic bransf drmational analysis o f . English sentences : An implement a t i o n . International Journal of Computer Mathematics, t o appear 3 . Sager, Naomi, and Grishman, Ralph. The r e s t r i c t i o n language f o r computer grammars of natural language. CACM",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "",
	"sec_num": null
	}
	],
	"back_matter": [],
	"bib_entries": {},
	"ref_entries": {
	"FIGREF0": {
	"uris": null,
	"type_str": "figure",
	"num": null,
	"text": "modificand i s o f t e n s t a t e d as a r e a s o n a b l e heuristic in n a t ur a l language parsing. (This i s e s s e n t i a l l y K i m b a l l ' s P r i n c ip l e Number S o * , \"Terminal s p b o l s o p t i m a l l y a s s o c i a t e t o t h e lowest nonterminal node.\" Although Kimball c a l l s this p r i nc i p l e \" r i g h t a s s o c i a t i o n \" and i l l u s t"
	},
	"FIGREF1": {
	"uris": null,
	"type_str": "figure",
	"num": null,
	"text": "\" m o d i f i e s \" o r d e r s \" .A \" p o t e n t i a l modificandrt must be a c c e p t a b l e pragmaticall'y as w e l l as s y n t a c t i c a l l y and s e m a n t i c a l l y .T h i s means t h a t i f t h e system r e f e r r e d t o i n t h e above example were s p r e a d Kimball, John. Seven*principles of surface structure p a r s i n g i n natural language. Cognition 2 (I), 15-47. 1 4 t h ACL Meeting 6 o u t over the e n t i r e country and j u s t had one of its processing stations in New York, the p r e f e r r e d analysis would have had \"in New York\" modifying \"processed1' instead of \"system\". Similarly, dialogue context can have an influence on what qualifies as a potential modificand. This paper describes an e a s i l y computed m e t r i c that involves a t t a c h i n g a number to each syntactic unit as it is formed, t o rank the possible analyses f o r any p o r t i o n of an utterance according to the above-stated heuristic. The technique is illustrated by its use in a system which supports a . n a t u r a 1 language d i a l o g u e f o r automatic business programming, This system u t i l i z e s syntax, semantics, pragmatics, and context to understand user utterances, 14th ACL Meeting MEDIAN S P L I T TREES: A FAST LOOKUP TECHNIQUE FOR FREQUENTLY O C C U R R I N G KEYS B. A. SHEIL HARVARD UNIVERSITY Median split (MS) trees are a new technique f o r searching s e t s of keys with highly skewed frequency d i s t r i b u t i o n s (such"
	},
	"FIGREF2": {
	"uris": null,
	"type_str": "figure",
	"num": null,
	"text": "depending on c o n t e x t and i n response t o the o p e r a t i o n s . The operations are described below with examples that show their importance in analyzing a l g o r i t h m d e s c r i p t i o n s : 1. Predicate interpretation: Inferences are sought to satisfy the demands predicates make on the n a t u r e of t h e i Among o t h e r t h i n g s , t h i s operation recovers o m i t t e d m a t e r i a l , f o r example when t h e whole i s used t o r e f e r to t h e p a r t . In \"T p o i n t s to a binary treew t h e p r e d i c a t e \"point\" r e q u i r e s i t s second argument t o be a node. The knowledge s t o r e d i n the d a t a base about b i n a r y trees i s searched f o r a dominant node, t h e r o o t node i s found, and the sentence becomes 'IT points t o the r o o t node of a binary tree. I t --2 . I n t e r s e n t e n c e r e l a t i o n s a r e determined by matching successive sentences a g a i n s t a small number of p a t t e r n s . These p a t t e r n s a r e s t a t e d i n terms of i n f e r e n r e s t o b e drawn from t h e c u r r e n t sentence and a previous sentence. A n o r d e r e d h e u r i s t i c search of t h e data base i s used t o l o c a t e t h e desired i n f e r e n c e s . Intersentence r e l a t i o n s a r e e s p e c i a l l y important i n algorithm d e s c r i p t i o n s , f o r they determine the flow of c o n t r o l i n t h e program. The most common and e a s i l y handled patterns in these texts a r e Temporal Succession, cause, and Enablement. They a r e t r a n s l a t e d i n t o successive l i n e s of code. But a l s o important i s t h e Contrast p a t t e r n : \"The predic a t e s of sentences S1 and S2 a r e t h e same. One p a i r of corresponding arguments a r e contradictory. The other p a i r a r e different: but s i m i l a r . ' Consider \"If VAL(M) VAL(N), s e t N equal t o L I N K ( N ) . I f VAL(M) < VAL(N)\\, s e t M equal t o LLNK(M). 11"
	},
	"FIGREF3": {
	"uris": null,
	"type_str": "figure",
	"num": null,
	"text": "f o r P r i m i t i v e s : Certain predicates a r e p r imitive'' tb t h i s a p p l i c a t i o n , i . e the t a s k component has s u b s t i t u t i o n r u l e s f o r them. Where p o s s i b l e , a decomposition of the t e x t i n terms of these p r i m i t i v e s i s looked f o r . For"
	},
	"FIGREF4": {
	"uris": null,
	"type_str": "figure",
	"num": null,
	"text": "Hobbs, Jerry R . A g e n e r a l system f o r semantic analxsis of English and its u s e in drawing maps Prom d i r e c t i o n s . AJCL Microfiche 3 2 , 1 9 7 5 . Hobbs,.J. R . , and Grishman, Ralph. The automatic t r a n s -Naomi, and Grishman, Ralph. The restriction language f o r computer grammars o f natural language. CACM 18 ( 7 ) , * J u l y 1975, p . 3 0 ."
	},
	"FIGREF5": {
	"uris": null,
	"type_str": "figure",
	"num": null,
	"text": "a t group of nouns, i n the object position; which cause the selection of the same sense for the verb in question, the common t h r e a d of meaning is extracted and posited as a semantic feature descriptive of that group of nouns. R e p e t i -t i o n of this process for various verbs and nouns yields a s e t of semantic features. These features are considered as f o rmalizing noun meaning at least f o r purposes of verb sense selection. The major problems in the above approach are 1) the i n e v itable intrusion of subj e c t i v i ty i n t~ the feature extraction task and 2) the lack of generality of the f e a t u r e s e t beyond t h e particular group of nouns and v e r b s s t u d i e d . Several methods have been used t o c o h t r 0 1 s u b j e c t i v i t y : r e p e t i t i o n of the process using the same verbs with different nouns and viceversa, and t h e generation of features from t h e regular study of other tasks of understanding which involve nouns--nounadjunct:noun pairs ( e . g . school house) and noun a n a l o g i e s . To ensure as much generality a s p o s s i b l e , t h e words considered i n t h e s t u d y have been chosen from the s e t of (statistically) most common English words and do n o t comprise any unified semantic f i e l d ."
	},
	"FIGREF6": {
	"uris": null,
	"type_str": "figure",
	"num": null,
	"text": "have been i n t e r e s t i n g and encouraging. In a d d it i o n t o t h e commonly used features (\"human, 11 I t animate, 1 1"
	},
	"FIGREF7": {
	"uris": null,
	"type_str": "figure",
	"num": null,
	"text": "c i f i c a l l y , t h e s e innovations include --Case a n a l y s i s , r a t h e r than f u n c t i o n a l d e s i g n .--Independent judgment of t h e adequacy-of t h e model.--Results i h the form of i n d i v i d u a l a l g o r i t h m s , n o t f u l l s y s t e m s . --Algorithms which a r e t r a n s f e r a b l e into nonresearch sys terns, W e d e s c r i b e a r e s e a r c h methodology, and t h e r e s u l t i n g Dialogue Modeling System which i n c o r p o r a t e s these innovations. The methodology analyzes n a t u r a l l y -o c c u r r i n g dialogues r a t h e r than prepared examples. For each dialogue we choose t o anal y z e , a s e p a r a t e model i s c o n s t r u c t e d . Each model i s judged a g a i n s t annotations of t h e dialogue prepared by a t r a i n e d obs e r v e r who uses a predefined set of c a t e g o r i e s of phenomena. Although t h e processes employed i n each model a r e t o b e as general as w e f i n d p r a c t i c a l , they may a l s o be a s ad-hoc a s necessary t o make t h e model work. After s e v e r a l models have been produced i n t h i s f a s h i o n , we expect t o f i n d t h a t , although many processes have been too i d i o s y n c r a t i c t o be used i n more than one model, many others have seen repeated application and thus have demonstrated t h e i r more g e n e r a l u t i l i t y . These proc e s s e s are regarded a s the p r i n c i p a l r e s u l t s , s i n c e they have been v e r i f i e d a g a i n s t a d i v e r s i t y of o b s e r v e r s ' a n n o t a t i o n s . Subsequent speakers w i l l provide d e t a i l s on t h e memory organization of t h e model, the processors which o p e r a t e on t h e s e memories, and t h e i s s u e s raised by t h e i n c l u s i o n of t h e observer i n t h e madel development c y c l e . 14th ACL Meeting KNOWLEDGE STRUCTURES F O R MODELING DIALOGUE This paper describes the know7 edge structures developed f o r our Dialogue Modeling System. Permanent knowledge is stored as Concepts in e semantic n e~w o r k Long Term Memory (LTM) , using a predicate/parametcr representational format Concepts t h a t are currently salient t o the Processors of t h e System are represented by tokens of those concepts, c a l l e d A c t i v a t i o n s , i n the Workspace (WS). A c t i v a t i o n s are transient entities, w i t h new ones frequently appearing and e x i s t i n g ones changing o r d i s a p p e a r i n g , The WS a t any moment forms the c u r r e n t con-~e x t , which drives c u r r e n t p r o c e s s i n g . A l l Processors intercommunicate through the WS, modifying the set of Activations and being influenced by them. The W S simplifies the integration of multiple sources of information about any given e n t i t y Two k l n d s of higher l e v e l knowledge s t r u c t u r e s are used t o model d i a l o g u e s . There a r e Execution Scenes, r e p r e s e n t i n g organized c l u s t e r s of a c t i o n s and/or objects being d i s c u s s e d in t h e dialogue'. These are s i m i l a r t o the o t h e r c u r r e n t l y"
	},
	"FIGREF8": {
	"uris": null,
	"type_str": "figure",
	"num": null,
	"text": "of dialogue, and a set of conventional subgoals f o r mutually achieving t h e i r goals.14th ACL MeetingI n our analysis of n a t u r a l l y occurring dialogues, we have* found many d i f f e r e n t kinds of Dialogue-games. including Helping C Smalltalk, Gripe, Polite-conversation, Order, Informationseeking, and Information-probing. An a n a l y s i s of fourteen helping dialogues uncovered a s u r p r i s i n g amount of consistency i n the s t r u c t u r e of these i n t e r a c t i o n s , which i s captured i n the representation of the Helping Dialogue-game.PROCESWRS F O R MODELING D I A L O G U EThis paper provides d e t a i l s of the major Processors incorporated i n t o our Dialogue Modeling System. These Processors a r e independent and asynchronous, monitoring the a c t i v i t y i n the Workspace, and modifying i t as appropriate. PROTEUS PROCESSOR --Every a c t i v a t i o n i n the Workspace has a \"level of a c t i v a t i o n \" (a number) which corresponds t o i t s degree of s a l i e n c e t o the model's c u r r e n t activity. Each of these activations imparts some of i t s s a l i e n c e t o its i m ed i a t e neighbors (including those i n Long Term Memory). S i m il a r l y , the a c t i v a t i o n l e v e l of each node i s augmented by t h a t imparted t o i t by a l l of i t s neighbors. Whenever a concept i n LTM accumulates a c t i v a t i o n above a certain t h r e s h o l d , an activ a t i o n of t h a t concept i s created i n the Workspace, Thus, if enough concepts, closely r e l a t e d t o X a r e active, X w i l l also become a c t i v e . Proteus i s the Processor which a t t e n d s t o a l l of t h i s , thus providing a mechanism f o r bringing a concept into attention, based on the a t t e n t i o n d i r e c t e d toward related concepts. 14th ACL Meeting MATCH PROCESSOR --This i d e n t i f i e s concepts i n LTM that are congruent to existing.activations in the WS The Dialogue Modelling System contains a number of equivalence-like relat i o n s . which Match uses to identify a concept i n LTM a s representing the same, t h i n g as an a c t i v a t i o n of some seeminglyd i f f e r e n t concept Once such an e q u i v a l e n t concept i s found, i t i s a c t i v a t e d . Match thus p r o v i d e s the same sort of a t t e n t i o n directi.cn that Proteus d o e s , except t h a t Proteus i s driven by the e x p l i c i t connectivity of the memory, w h i l e Match responds t o more indirectly-represented s i m i l a r i t i e s . DEDUCE PROCESSOR --T h i s attempts t o apply a r u l e , whenever ehat r u l e Bas become-a c t i v e . Rules are concepts of the f o r m . \"(Conditior)) -> (Action) \" . They become a c t i v e by v i r t u e of the a c t i v i t y of the concept which i s t h e i r condition h a l f . Deduce *I senses t h e a c t i v i t y of a r u l e and attempts t o a p p l y i t by a c t ivating the concept f o r t h e a c t i o n . Whateve'r correspondences were evolved i n t h e course of creating t h e a c t i v a t i o n of t h e condition (left) half of the rule are c a r r i e d over into t h e a c t i v a t i o n of t h e a c t i o n (right) h a l f . The combination of Match and Deduce gives us a l l the c a p a b i l i t y o f a production system, where the system i s represented by a l l the Rules i n LTM. DIALOGUE-GAMES PROCESSOR --Dialogue-games a r e a theoretical construct w e have developed to represent c e r t a i n c o m u n icative f o m s used by people t o i n t e r a c t i n achieving goals each person has. Each Dialogue-game has a s e t of parameters, including two roles and a t o p i c . Each parameter has a s e t of 14th ACL Meeting specifications on what can serve as a value f o r itself. We"
	},
	"FIGREF9": {
	"uris": null,
	"type_str": "figure",
	"num": null,
	"text": "Dialogue-game .PRONOUN PROCESSORS --The Dialogue Model System contains a set of Pronoun processors, including an I-Processor, a You-Processor, and an It-Processor. Each of these is invoked whenever the associated surface word appears in an input utterance, and operates to i d e n t i f y some preexisting activation that can be seen as r e f e r r i n g t a the same o b j e c t .OBSERVATION METHODS FOR MODELING DIALOGUEMost models of natural language are constructed using only the designer's own linguistic intuitions. and'evaluated by others using t h e i r own intuitions. This paper describes observation methods which have been used as a guide for the generation of process models of dialogue phenomena and as a basis f o r e m p i r i c a l evaluation of these models.The methods have concentrated on a number of specific dialogue phenomena. Observers w e r e given the annotation instructions, and after some training, produced highly reliable 14th ACL Meeting annotations of :Requests -Annotations of requests and the various kinds of responses to these requcsts by t h e e t h e r p n r t i c ipant in a dialogue were useful for evaluating various control issues.Repeated reference -Annotations of repeated r e f e r e n c e were used to evaluate models of anaphoric reference (cases in which two sets o f words in a dialogue describe the same thing). Topic -Observers' anriotations of the topic units in naturally-occurring dialogues provide a basis for specifying high level multi-utterance knowledge structures. Expressions of comprehension -Observers annotated t h e expressions of comprehension (or l a c k thereof) by each participant of the other's previous utterances. These annotations a r e being used t o evaluate our model of the stopping rule for comprehension, which determines when the model has successfully comprehended an utterance. Similar expressions -Observers judged possible paraphrases of the utterances of dialogues for similarity in meaning in isolation and also in the context of the dialogue. This is serving as useful data for determining the effect of context on comprehension. 14th ACL Meeting Correction actions -Observers' annotations of utterances t h a t corrected some information conveyed previously i n the dialogue allow us to evaluate our representation of the comprehension of the previous utterances. While these phenomena don't span a l l possible i s s u e s of i n t e r e s t in natural language comprehension, t h i s same g e n e r a l methodology can be ext-ended to other 1anguage.phenomena. It can provide a valuable guide for building n a t u r a l language models, and an empirical basis for evaluating models. 1 4 t h ACL Meeting A PROCESS MODEL FOR SPEECH ACT UNDERSTANDING In a canversation, hearers must make judgments about the i n t e n t i o n of a s p e a k e r ' s utterance. A restaurant waiter r e alizes t h a t when a patron says, \"Waiter, t h i s soup i s c o l d , \" he is not merely thereby making an a s s e r t i o n , b u t is a l s o making a complaint, and perhaps requesting r e c~i f i c a t i o n . Utterances can be understood t o hav'e more than one i n t e nt i o n . Two types of kntentions are described i n t h i s p a p e r , both of which depend l a r g e l y on c o n t e x t u a l f a c t o r s for t h e i r r e c o g n i t i o n o r understanding i n a dialogue s i t u a t i o n . Those of t h e f i r s t type (type A) are broadly a p p l i c a b l e t o many d i s c o u r s e s i t u a t i o n s and are understood p r i m a r i l y on the b a s i s of t h e syntactic s t r u c t u r e o f utterances and certain f a c t s about t h e s t a t u s r e l a t i o n s h i p s that hold between t h e speaker and the hearer. Type A i n t e n t i o n s are very similar to the Speech Acts of Searle o r of Gordon and Lakoff. Speech a c t s of t h e second type (type B) a r e much more dependent on t h e nature of t h e p a r t i c u l a r contextual situation. They. are recognized o r understood on t h e b a s i s of shared cultural norms concerning the p a r t i c u l a r kind of conversation the p a r t i c i p a n t s a r e engaged i n . Type B speech a c t s a r e much l e s s general in n a t u r e than type A . I f a salesman says, \"This l i t t l e two-tone model has been p r i c e d t o s e l l , \" then t h e customer should recognize that the type A intention is to make an assertion or a claim, and that the type B i n t e n t i o n i s t o o f f e r an i t e m f o r s a l e . 14th ACL Meeting A prdcessing model for speech act recognition or understanding is presented, in wMch procedures c a l l e d spe-ech a c t schemata are activated. In the example with the salesman, the customer should have experienced the a c t i v a t i o n of both an assertion schema and a sales-offer schema The schemata are formulated i n SOL, the high l e v e l language of the LNR memory model orma man, Rumelhart, and LNR), which permits t h e modeling of conceptual e n t l t i e s which have both s t r u c t u r a l and procedural p r o p e r t i e s . I n a d d i t i o n , the t y p e B speech a c t schemata are subprocedures of larger information o r g a n i z e r s which a r e sensitive t o the p a r t i c u l a r c o n t e x t . These l a r g e r devices are called scripts and are r e l a t e d t o the computational e n t it i e s of t h a t name developed by Schank and Abelson. In the example with the salesman, t h e s a l e s -o f f e r speech a c t i s p a r t of the \"Commercial Transaction\" s c r i p t .A detailed model for one such script, an \"Airline Reservation Phone Call\" s c r i p t i s presented, together with an account of the i n t e r a c t i o n s between s c r i p t s , t h e two types of speech a c t schemata, and the syntactic s t r u c t u r e of u t t e r a n c e s ."
	},
	"FIGREF10": {
	"uris": null,
	"type_str": "figure",
	"num": null,
	"text": "i t s goal the unification of current ideas about problem solving w i t h current ideas about conteatual language comprehension. To this end, the CSA project currently consists of e k r e e gart's, all built on top of the declarative CSA representation formalism: (1) a plan synthesizer, (2) a sentencein-context interpreter (designed as the 2-sentence case of an eventual n-sentence story comprehender), and (3) a mechanism descriptllon \"laboratoryr', who~se purpose is to provide a framework &or describing the \"causal topology\" of man-made devices and mechanisms. The LISP sys tern -which implements t h e"
	},
	"FIGREF11": {
	"uris": null,
	"type_str": "figure",
	"num": null,
	"text": "s t o r y , The Magic Grinder (a Walt Disney book-of-the-month book), which the CSA group is employing to focus the language 14th ACL Meeting comprehension part of t h e p r o j e c t . It is believed t h a t any language comprehender must possess CSA-like knowledge, and t h a t any theory of language comprehension must d e a l w i t h issues quite similar t o those encountered by plan synthe-This paper describes a computational approach t o t h e problems of discourse s t r u c t u r e of real-world English paragraphs, u s i n g the techniques G \u00a3 h e u r i s t i c search. Procedures for detecting intersentence relations a r e being developed within the framework of a system f o r semantic analysis o f , English texts (1). This system takes a s input the t e x t i n a predicate c a l c u l u s -l i k e notation produced by a s y n t a c t i c f r o n t end ( 3 , 2 ) . Various s e m a~t i c operations a r e applied: 1) Inferences are sought t o s a t i s f y the demants made by predicates on the nature of their arguments ( p r e d i c a t e i n t e rp r e t a t i o n ) ."
	},
	"FIGREF12": {
	"uris": null,
	"type_str": "figure",
	"num": null,
	"text": "Intersentence r e l a t i o n s are determined by matching the current sentence and the previous text against a s m a l l number of p a t t e r n s . These patterns are stated i n terms of inferences t o be drawn from the c u r r e n t sentence and an \" e l i g i b l e \" pre-vious sentence, and t h e modification t o be performed on the text i f the p a t t e r n i s matched. The patterns a r e of t w o 14th ACL Meeting kinds--coordinating and subordinating. One coordinating relation is Temporal Succession : \"The current sentence asserts a change whose initial s t a t e i s implied by an e l i g i b l e previous sentence.\" Other coordinating relations are Cause, Contrast, Parallel, and Paraphrase. Two subordinating relations are 1) Example: \"The predicate and arguments of the assertlon of the current: sentence stand in a subset or element-of relation to those of the assertion of ah eligible previous sentence. \" (See below.)2) \"Relative clause in disguise\": \"The current sentence provides further information about an entity fairly deeply embedded in an eligible previous sentence. \" (E.g. the second and third sentences of this abstract.) When this pattern is matched, the modiffcation made to the text is equivalent to. turning the current sentence into a relative clause."
	},
	"FIGREF13": {
	"uris": null,
	"type_str": "figure",
	"num": null,
	"text": "Hobbs, J e r r y R . A general system for semantic anal y s i s of E n g l i s h and i t s use in drawing maps from d i r e c t i o n s . AJCL Microfiche 3 2 , 1975.-American Journal of Computational Linguistics ~i c r o f i c h e 51 : 39.ACM CWPUTER SCI ENCE CONFERENCE E M P L O Y M E N T R E G I S T E RThe Fifth Annual R e g i s t e r w i l l provide books of l i s t i h g s of a p p l i c a n t s and p o s i t i o n s during the Conference .Applicant listings fnclude education, p u b l i c a t i o n s , experience, interests, r e f e r e n c e s , p o s i t i o n , and salary d e s i r e d . Employer listlngs include p o s i t i o n available, s t a r t i n g date, salary, and benefits; education, experience, and s p e c i a l i z a t i o n o r a p p l i c a n t s , plus $5 f o r anonymous listing Free to students A form on which t o submit i n f o m a t i b n w i l l be supplied by American Journal of Computational Linguistics Microfiche 51 : 46 L A T S E C : A supplementary payment of $400,000 on an indexing c o n t r a c t was criticized on the floor of the House by Congressman Aspin of Wiscondin. According t o t h e Congressional Record (May 4 , H3909), Aspin s a i d t h a t t h e payment was made \"under the provisions of a special b a i l o u t l a w (Public Law 85-804)\" when Peter Toma \"said that he would be 'impelled t o leave t h e United States' without the a i d . \" The b a i l o u t law, according t o Aspin, \"should only b e used when the Pentagon r e a l l y needs equipment that i s v i t a l t o the n a t i o n a l s e c u r i t y . \" The Latsec c o n t r a c t was one of s e v e r a l t h a t Aspin characterized as not q u a l i f y i n g f o r t h e supplementary payments. He c a l l e d f o r a new stady of t h e whole concept. Torna was quoted i n newspaper r e p o r t s ae* saying t h a t \"If Conkressman-Asgin were t o look a t our system and what i t i s doing for t h i s country, he would withdraw his c r i t i c i s m . \" The newspaper s t o r y describes the b a s i c c o n t r a c t as one w i t h t h e Foreign Teohnology Division*.at Wright-Patterson A i r Force Base, originally worth $460,000 f o r a three-year p e r i o d ."
	},
	"FIGREF14": {
	"uris": null,
	"type_str": "figure",
	"num": null,
	"text": "merely practical and lacking in theoretical interest. It is therefore i r o n i c that had more research been done directly on translation, the development of linguistic theory itself might have been accelerated b y d i v e to ten years. (Interestingly, transformational linguists, so often l i n k e d w i t h computers i n the publi'c mind, had little involvement wich MT.)"
	},
	"FIGREF15": {
	"uris": null,
	"type_str": "figure",
	"num": null,
	"text": "one of the flew p r o j e c t s in recent years to t r y t h i s , at Berkeley, was curtailed last year f o r lack o f f u n d s . )"
	},
	"FIGREF16": {
	"uris": null,
	"type_str": "figure",
	"num": null,
	"text": "Microfiche 51 : 50 I I A C H I N E T R A l i S L A T I O N : Moscow INTERNATIONAL SEMINAR Head, Machine Translation Department All-Union Centre for Translation of Scient'ific and Technical Literature and Documentation U1. Rrzhizhanovskogo 14, korp. 1 117218 Noscow Bb218 An Iqternational Seminar on Machine Translation took place in Moscow from 25 to 27 November, 1975. About 200 scientists from People's Republic of Bulgaria, the Getman Democratic Republic, Czechoslovak Socialist Republic, People's Republic of. Poland, and the Soviet. Union participated in the Seminar which was organized by the All-Union Center for Translation of Scientific and Technical Literature and Documentation. In his opening speech the head of the All-Union Center for Translation (ACT), Dr. V. N. Gerasimov, said that the expansion of interational contacts and inte~nationalization of science. had promoted the growth of the translation activities in the USSR. In the country at present translation is being done by various specialized and departmental organizations. In 1975, for example, the ACT translated more than 30,000 author's sheets of scientific and technical literature and documentation. In the MTI Moscow INTERNATIONAL SEMINAR nearest future ACTgalone will reach the volume of 50-80,000 printer's sheets a year. Dr. Gerasimov sees the way out of this situation in the opeedies t possible deveLopment and installation of industrial systems of machine translation. This was also the canclusion of the Temporary Scientific and Technical Commisslton on Machine Translation of the State Commission on Machine Translation of the State Committee on Science and Technollogy of the Council of Ministers o f the USSR which worked in 1972-1973 At the plenary session, four reports dealing with general principles of construction of machine translation systems and ways of Improvement of MT quality were delivered. Yu. N. Marchuk (ACT, Moscow) draws our attention to the increased role of computer dictionaries in automatic processing systems and namely in M systems. Since the quality of MT largely depends on the word-list and volume of the dictionary and the information contained in the dictionary entries, these dictionaries must, according to Marchuk, be compiled with due regard to distributional and statistical methods. It is important to make wide use of contextual information. Interand post'editing are absolutely indispensable in running the firsteindustrial MT sys terns. This idea is developed in a joint report by B. D. Tikhomirov, Yu. N. Marchuk, and I. I. Oubine (ACT, Moscow), who put forward a new approach to inter-and post-editing for correction of input and translation errors. Interediting follows a search in the MTt Moscow INTERNATIONAL SEMINARdictioriary, automatic a n a l y s i s of new words and formation of the information required f o r future processing. The intereditor e i t h e r confirms o r corrects the grammatical information of new word8 and can also c o r r e c t input e r r o r s , supply t r a n s l a t i o n s f o r new words, e t c . A t t h e i n t e r e d i t o r ' s command the system accumulates new words f o r the p o s t e d i t o r . The p o s t e d i t o r exercises phrase-by-phrase c o n t r o l of the t r a n s l a t i o n . I f t h e t r a n s l a t i o n of a phrase does not satisfy him he edits i t with t h e help of a display unit and, i f necessary, sends t h e o r i g i n a l phrase and i t s machine t r a n s l a t i o n t o a s p e c i a l storage device f o r f u r t h e r analysis aimed a t the improvement of t h e E1T algorithm: In h i s summarizing r e p o r t , D r . A . Ljudskanov CPRB) singled out the main stages of MT development, and showed what an unfavorable e f f e c t t h e ALPAC r e p o r t and t h e pessimistic statements of D r . Y. Bar-Hillel had produced on Em. tlT systems aimed t o operate on an i n d u s t r i a l s c a l e i n the nearest f u t u r e a r e t o be developed on t h e b a s i s o f the \" s e l e c t i v e f 1 s t r a t e g y , i.e. they a r e t o use only that information \"which i s necessary and s u f f ic i e n t f o r t h e given aim f o r the given p a i r of languages. I I Dwelling upon t h i s thought, D r . Ljudskanov s a i d i n h i s second r e p o r t , \"Lexeme d i c t i o n a r y f o r M systems\", that i n the MT s y stem which i s under development i n Bulgaria \"deep\" d i f f i c u l t i e s are s h i f t e d t o \"surfaceff l e v e l s ( l e x i c a l and morphological) and"
	},
	"FIGREF17": {
	"uris": null,
	"type_str": "figure",
	"num": null,
	"text": "s y s t e m s . Dr. Piotrovsky considers that vocabulary contains the lion's share of slntaqtic and semantic information of the text and consequently the basis of any MT system must be a bilingual automatic dictionary with the output word list of the thesaurus type: Such MT systems can be developed within reasonably short periods of time and will, according t6 Dr. Piotrovsky, enable the consumer to derive all semantic information from the input t e x t . Thus the main speakers p u t forward a s t h e caxdinal t a s k of the present day the creation of MT systems not yet oriented to high-quality translation but working industrially More than 60 reports were made at the seminar which were distributed into four sections: ('a) computer dictionaries; (b) aut~matic analysis and synthesis of texts; (c) semantic analysis of texts; (d) mathematical and program maintenance.Various trends in the theory and practice of machine translation were represented at the section on computer dictionaries."
	},
	"FIGREF18": {
	"uris": null,
	"type_str": "figure",
	"num": null,
	"text": "The c e n t r a l component of this Anglo-Russian Automatic Translation sys tern (ARAT) is a dictionary of a special type (Ailgla-Russian Multiaspect Automatic Dictionary--ARMAD) in which every entry gets complete and diverse characteristic properties of its linguistic behavior on all levels of d e s c r i p t i o n : morphological, lexical, syntactic, and semantic. For formal recording of a l l required infomition-, ARMAD employs syntactic p a t t e r n s , lexical functions, demands imposed by the word on the linear and structural context, rules of standard and individual modifications of lexical and syntactic structures with this word, fomal semantic representation, semantic selection of syntactic vglencies of the key word, etc. The abundance and variety of information contained in the dkctionary entry of practically eumy word is the principal advantage of ARMAD as compared a i t h other types of automatic dictionaries. Wide use of syntactic and expecially semartic information makes it p o s s i b l e w i t h the help of this dictionary in the framework of the ARAT system to solve such a cardinal problem of MT as grammatical and lexical ambiguities. Data on l e x i c a l cooccurrence of the original English word and its Russian equkvalent as well as on required syntactic transformations. it is hoped, w i l l ensure an exact and idiomatic translatibn of t h e original t e x t But at the same time the abundance and variety of thr information in ARMAD naturally leads to a more complicated entry structure and multiplies the difficulties which linguists have to overcome when writing entries. At present the d i f f i c u l t i e s and possibilities of algorithmic r e a l i z a t i o n of Pull-scale dictionaries of this type have not been investigated. A number of reports were devoted to compiling various types of auxiliary dictionaries and c r e a t i o n of whole systems for automatic lexicographic work. In their reports, Yu. N. Marchuk, N. G. Tikhonova, and I. I. Oubine ( a l l of ACT, Moscow) inforrhed the participants of the seminar of the work now being done to develop a bilingual automatic lexicographic system to compile frequency and semantic frequency d i c t i o n a r i e s and bilingual concordances as auxiliary material for E n dictionaries.E. V. ~ertei and V. A. Vertel ( S t a t e Institute of ForeighLanguages, Minsk) submitted a joint report on the elaboration of an a l g o r i t h andd,a seroof programs to compile a frequenc.y-alphabetic dictionary and concordance on a medium-size computer. The s e t of programs is designed to process an original text of up to 300,000 words.An interesting report on the compilation of a xcverse French dictionary was submitted by E. L. Kozmina (Applied Mothematics Institute, Moscow), who worked out a method of g e t t i n g a wordform by its ultima or by one wordform representing a whole class of"
	},
	"FIGREF19": {
	"uris": null,
	"type_str": "figure",
	"num": null,
	"text": "The section on automatic analysis and synthesis of texts attracted the largest number of reports, reflecting the intense interest of linguists in these important parts of MT systems.A series of interconnected r e p o r t s was read by scientific workers of the Computational Center of Leningrad State University Dr. MTI Moscow INTERNATIONAL SEMINAR G. S. Tseitin proposes t o use f o r automatic syntactic analysis models which do not simply contain a s e t of r u l e s f o r construct i o n (or \" f i l t e r i n g \" ) of admissible syntactic structures but a l s o mark out among t h e s e r u l e s more or less \"preferable\" (\"normal\", \"nuclear\", \"productivet') . B . M . Leikina (Leningrad S t a t e University) i s working on a model. of an English grammar which permits under c e r t a i n l i m i t a t i o n s nonprojectivity while f u l f i l ling the p r o j e c t i v i t y demands f o r the majority of cases. A t present this group i s working on a more complicated model which takes i n t o account among other things t h e order of establishment of s y n t a c t i c l i n k s i n a p a r t i c u l a r s t r u c t u r e ' A number of experiments have been c a r r i e d o u t with t h e first v a r i a n t of this g r a m a r . These experiments were aimed a t checking t h e presence of t h e c o r r e c t a n a l y s i s and absence of a fixed i n c o r r e c t a n a l y s i s and were carried out under t h e conditions of man-machine i n t e r a c t i o n during which a p a r t of t h e intermediate r e s u l t s were rejected by t h e man. Complete a n a l y s i s has been t e s t e d only on s e v e r a l sentences. The p r i n c i p l e s of a n a l y s i s i n the ARAT system were expounded i n t h i s * s e c t i o n . Besides t h e r e p o r t of Z . M. Shaljagina a l r e a d y mentioned, t h r e e r e p o r t s on less general problems were read: L. A . Afonasjeva, \"Disambiguation of homonymy i n che ARAT system:', 'P. N . Nikanorova, \"Prepositions i n the ARAT system\", and 0. A .Sternova, On a model of Russian i n f l e x i o n i n t h e ARAT system\" (all State I n s t i t u t e of Foreign Languages, Moscow"
	},
	"FIGREF20": {
	"uris": null,
	"type_str": "figure",
	"num": null,
	"text": "choice of the tree coryesponding to the lexical units of thephrase is made on the basis of t h e information contained i n the automqtic dfctionary. The said d e s c r i p t i o n is being done for the analysis of English scientific and technical t e x t s .The j o i n t r e p o r t of E. Benesova, A.. Bemova, S . Machova, and J. Panevova (Czechoslovakia) contains characteristics of cardinal components and principles of functional generative description designed to get semantic representations which in heir t u r n are t h e b a s i s f o r synthesis of sentences of a n a t u r a l language. Other reports a t the section on automatic analysis and synthesis of t e x t s were of a l e s s general c h a r a c t e r . D r . Klimonov (GDR) presented a s e t of c r i t e r i a for automatic ident i f i c a t i o n of antecedents of personal and possessive pronouns in' Russian and German. Dr. S tarke (GDR) read a report on a method of transformation of Russian structures with German equivalents having d i f f e r e n t dependency representations. Each transformation jrs performed by means of elementary operations: insertion, d e l e t i o n , a l t e r a t i o n of a node, e t c ."
	},
	"FIGREF21": {
	"uris": null,
	"type_str": "figure",
	"num": null,
	"text": "A number of reports show t h e i r a u t h o r ' s e f f o r t s t o solve t h e i r problems within the framework of automatic t e x t processing. These r e p o r t s are the d e f i n i t i o n of the concept of answer i n question-answer s t r u c t u r e s of texts ( D r . Konrad, GDR); semantics of prepositions (Leontjeva, ACT, Moscow) arci Nikitina ( I n s t i t u t e of Linguistics, Academy of MTI Moscow INTERNATIONAL SEMINAR 6 Sciences, Moscow); semantic and syntactic analysis of Rus~ian words with the meaning \"quantity\" (I1 j in and Smirnova , Leningrad State University) ; communicative structure of the 'Engltdh senteace (Korolev, ACT, Moscow) ; semantic analysis of headiqgs of Japanese patents (~einaiovk, .Samari~a, and Shevenko, Insf itute of oriental Studies, Academy of Sciences, Moscaw) . An interesting report was submitted by N. A. Kuzemskaya and Dr. E. F. Skorokhodko (Institute o f Cybernetics, Kiev), who reported work on quantitative criteria for estimation of translation quality. They propose to solve this problem with the help of a semantic language which would meet the following conditions: (a) a b i l i t y to express the required information; (b) possibility of getting the necessary quantitative estimates; (c) possibility of translating into this semantic language from t h~ original language and from t?he language of translation. Semantic networks can be used as such a semantic language. Comparison of two speech semantic networks--the primary corresponding to the original text and the secondary to the text. of the translation--makes it possible to define a number of parameters characterizing various aspects of the quality of the translation. The main aspects of translation quality are completeness and accuracy. Completeness is c a l c u l a t e d from t h e proportion of the primary network contained in the second, and accuracy from the proportion of the secondary that coincides with the primary. The problem of automatic recognition of key words in text was t h e main concern of the reports of R . A . Kovalevitch, V. A. Sorkina (State Institute of Foreign Languages, Minsk), and G. S Osipov, A ; I. Chaplj a (MaQachkala State University) . Kovalevitch and Sorkina singled out more than 50 elementary semantic units and, relying on formal characteristics o f the text and using distributiorial statisticdl methods, they singled out and formalized rules of combination of elementary semantic units in free word-groups. The result of the analysis i s a matrix of relations of components in English word groups and their Russian translations with all necessary grammatical information. The set of formal characteristics of all components of a free English word group determines the corresponding Russian equivalent. In their report, Osipov and Chaplja spoke about an attempt to construct (by the thesaurus method) formal means useful. for elaboration of an algorithm of recognition of key words in a text. These authors build this formal apparatus on the basis of estimating he degree of synonymy of words in the structures of the input text. To our regret it should be admitted that i n spite of the universally recognized importance of the semantic component for MT, the seminar lacked reports in which authors put fiorward methods and instrments of formal representation of word meanings for' the solution of problems of analysis and synthesis of texts during MT. The participants of the seminar did not pay sufficient attention to the search f o r optimum ways of elaborating formal semantic languages for MT. . MTt Moscow INTERNATIONAL SEMINAR Ip the software section, tbn reports wexe made, three of which are characterized by a broad approach t o this important part of flT systemg, These are the r e p o r t s of N. A . K t~p k o and G. S. Tseitin (Leningrad S t a t e University), V. S. Krisevitch and I. V. Sovpel (Institute of Foreign Languages, Minsk), and D. M. Skitnevsky (Institute of Foreign Languages, Irkutsk) The j o i n t r e p o r t by Krupko and Tseitin sums up the research work carried out during the last 13 years, involving development of software to be used in computer experiments on MT. At present the software of the MT model at the MT laboratory of Leningrad State University comprises the following components : (a) the system of*symbolic representation of the linguistic content of the model; (b) the machine representation of t h e linguistic content ; (c) the compiler for translating the symbolic representation into the machine representation; (d) the interpreter which employs t h e machine representation t o process a given t e x t i n accordance w i t h the aim of the experiment. Such a structure of the software is motivated mostly by the necessity of changing the linguistic content in the course of experiments and by the impossiblity of keeping in core memory more complicated linguistic information in the form of a compiled program. The same approach with an appropriate shift in the structural role o f each of the component parts seems, in the opinion of the authors of the report, to be quite reasonable in developi~g practical Fp systems. MTt Moscow INTERNATIONAL S E M I N A R 64 The r e p o r t by Skitnevsky considers some basic principles bf software develbprnent for linguistic investigations. The main idea of Skitnevsky's model is t o consider 8s interdependent the following three levels of software : (a) the conceptual level-a model o f linguistic process; (b) the compositional level--an input languake; (c) the programming level--the implementation system,. The development of software includes the following five stages: the f i r s t stage i s an analysis of a s e t of algorithms which a r e given i n t h e i r verbal d e s c r i p t i o n ( i . e . a s f l o w c h a r t s ) . The analysis results in an informal d e f i n i t i o n of the class of algorithms. The second stage consists in s~e c i f y i n g this formal definition. As a result we have a l o g i c a l and mathematical framework, which i s a sufficiently precise and dynamic descript i o n o f basic concepts occurring in the informal d e f i n i t i o n (MODEC). A t the t h i r d s t a g e a d a t a bank and a service program package using the standard software a r e formed on the basis of the set of o b j e c t s and operations of,the model. The f o u r t h stage c o n s i s t s of working o u t an input language which permits tbe user t o s t a t e h i s tasks i n terms o f t h e MODEL. The f i f t h stage designs an o p e r a t i o n a l system which uges the standard software and remains i n t a c t i n respect t o relatively extend-ed MODEL, its data bank and service program package. The MODEL built along these l i n e s and an input language form a convenient metalanguage f o r the description of the computer semantics of natural languages. The report by Krisevitch and Sovpel emphasizes that an MT system operates with large information f i l e s including input and"
	},
	"TABREF0": {
	"html": null,
	"type_str": "table",
	"content": "<table><tr><td>14th ACL Meeting 14th ACL Meeting</td></tr><tr><td>within the application part the node for the concept QUANSHIPD</td></tr><tr><td>has a SUPerset arc to QUANT and a KIND a r c t o 1 . Hctwccn I N THE P H L I Q A 1 SYSTEM</td></tr><tr><td>parts of the network, QUANSHIPD is linked as t h e M O (Relilted</td></tr><tr><td>A p p l i c a t i o n Object) arc of the program node QS, the language</td></tr><tr><td>node QUANTIT (the stem of \"quantity\") is linkcd as the RLO A large number of the most commonly used,English nouns</td></tr><tr><td>(Related Language Object) of QUANT, and t h e 1angua);c node SHIP belongs to the category of mass nouns. In recent years it</td></tr><tr><td>i s the RLO of SHP. I n g e n e r a l , the links bet~,*\\.ecn cliljnccnt has become increasingly clear that the semantic analysis of</td></tr><tr><td>p a r t s are many-to-many. expressions containing mass nouns encounters great difficulties.</td></tr><tr><td>The linguistic processing in our system is specified by because of the uncertain logical status of mass noun referents.</td></tr><tr><td>augmented phrase structure rules, uniformly for the three</td></tr><tr><td>levels of processing: semantic, syntactic, and morphological.</td></tr><tr><td>Given a p o i n t e r to a node in the program part, a l~n g with some</td></tr><tr><td>information about what aspect of that node is to be explained,</td></tr><tr><td>these rules can generate an appropriate English statement by</td></tr><tr><td>utilizing the ~tructure of the network. Because these s t n t e -</td></tr><tr><td>ments are constructed from very small p i e c e s ( i . e . word stems),</td></tr><tr><td>considerable flexibility of expression is possible, including</td></tr><tr><td>taking into account dialogue context.</td></tr><tr><td>This paper will describe in some detail the overall</td></tr><tr><td>method of generation, with special emphasis on complex definite</td></tr><tr><td>mantic network with three p a r t s . The program part contains</td></tr><tr><td>information about the programing language and about a specific noun phrases.</td></tr><tr><td>program; the application part contains information about business</td></tr><tr><td>concepts; and the language part contains information about English</td></tr><tr><td>words. Each of these parts has its own internal structure, and</td></tr><tr><td>there are links between the parts where appropriate. For example,</td></tr></table>",
	"num": null,
	"text": "In our system, information is stored in the form of a se-SEMANTIC INTERPRETATION OF MASS NOUN EXPRESSIONS"
	},
	"TABREF1": {
	"html": null,
	"type_str": "table",
	"content": "<table><tr><td>14th ACL Meeting</td></tr><tr><td>a theory, i n which a ceritral role is played by a l o g i c a l formalism, developed for the semantic representation of mass A c e n t r a l problem for computational models of understand-Medema, f . , W. P e l l e t i e r , F . J . , ed. On mass terms. Synthese 31 ( 3 / 4 ) , . 1 9 7 5 . ing is to formalize the meanings of words which are manipulated</td></tr><tr><td>expressions. These r e p r e s e n t a t i o n s were required t o meet t h e Scha, R . J. H. Semantic types i n PHLIQA 1. Preprints, by the model. This paper reports on research directed towards</td></tr><tr><td>following conditions: COLING 7 6 * the formalization of the meanings of nouns. The g o a l of t h e</td></tr><tr><td>Winagrad, T. Understanding natural language. Edinburgh research is to d e e c r i b e nouns sufficiently t o enable the per-</td></tr><tr><td>formance of a particular task of understanding: the appropriate</td></tr><tr><td>University Press, 1972.</td></tr><tr><td>selection of verb senses for multiply-sensed verbs.</td></tr><tr><td>at present is grdssly Work in computational linguistics so far has not contri-underdeveloped\" (McCawley 197 5) buted to a solution or even a better understanding of the problems that mass nouns pose. In language understanding sys-tems that have been developed in the past few years these problems have been evaded, either by restricting the admitted use of mass nouns trivially, as in the SHRDLU dialogue system (Winograd 1972), or by explicitly excluding their use altoge-ther, as in the LSNLIS question answering system (Woods 1972) . The recently implemented question answering system PHLIQA 1 (~andsbergen 1976, Medema et al. 1975, Scha 1976) ; i . e . no articulation of the r e f e r e n t n o t to The hypothesis being tested is that noun meaning can be Woods, W. R e p q r t , d e s c r i b e d The familiar; the contributions of this work are the development of empirical methods of analysis and the generation of a general set af features. The effect of noun meaning on the sense of a verb can b e clearly seen by examining simple sentences with identical syn-tactic structure and no extra-sentential context. Tb study noun meaning in a regular, non-subjective fashion, a large number of simple sentences of the form noun-verb-noun have say that the r e f e r e n t Lexicography and t h e count-mass d i s t i n c t i o n . been generated.</td></tr><tr><td>may be considered as a su~srantial step forward in this Proceedings, Berkeley Linguistic S o c i e t y , 1975.</td></tr></table>",
	"num": null,
	"text": "subject (seePelletier 1975) . It is indeed fair to say that, \"the logic of mass expressions . . . by primitive \"units of meaning, \" here called features . concepts of semantic features and componential analysis are . In groups of the sentences, the subject noun and verb have been fixed, while the object noun is varied."
	},
	"TABREF3": {
	"html": null,
	"type_str": "table",
	"content": "<table><tr><td>syntac-</td></tr><tr><td>tically aided comprehension\" procedure is used; cogn-itive</td></tr><tr><td>processes, taking advantage of the uniform representation of</td></tr><tr><td>knowledge, utilize linguistic knowledge as is required< for</td></tr><tr><td>tasks such as the disambiguation of conceptual relationships,</td></tr><tr><td>and the location, in the network, of referents designated by</td></tr><tr><td>phrases instead of names. Altheugh the system has the syntactic</td></tr><tr><td>knowledge needed to</td></tr></table>",
	"num": null,
	"text": "14th ACL MeetingNATURAL LANGUAGE UNDERSTANDING BY C O G N I T I V E NETWORKSA model of human cognition which stores and utilizes l i nguis tic knowledge in the same structures as 'general ' or 'world 'knowledge. supports an efficient and tntuitively satisfying technique for the understanding of natural language. Cognitive comprehension and syntactic passing can proceed in parallel, cooperatively, using the same procedures, without the loss of the syntax/semantics disrinc~ion ."
	},
	"TABREF4": {
	"html": null,
	"type_str": "table",
	"content": "<table><tr><td>THE FUTURE OF MT</td><td>48</td></tr><tr><td>support f o r</td><td/></tr><tr><td colspan=\"2\">The 1966 report of the Automatic Language Processing Advisory</td></tr><tr><td colspan=\"2\">Committee (ALPAC), which concluded that MT results had not been</td></tr><tr><td colspan=\"2\">fully satisfactory, led to the virtual elimination of government</td></tr></table>",
	"num": null,
	"text": "is a widespread myth among linguists that machine translation+-or, properly, machine-aided translation--which was the object of intense effort and research a decade and a half ago, was found to be a failure and has since been abandoned.Nothing, in fact, could be further from the truth.Although a number of institutions and agencies in the U. S. The Georgetown program was the ultimate basis for two of the major functioning MT programs in the U.S. today, that at Oak Ridge and at ~ri~ht-patterson Air Force Base.. These and other programs every year produce thousands of usable scientific and technical translations. However, they are all built on a research base which is now nearly twenty years old. MT research. While the conclusion was not strictly justified (for example, scientists at Oak Ridge and Euratom,, given choice between human and machine translation, both opted for the latter), the reduction in funding was timely, since the extant programs had largely exhausted the then-available possibilities in computer technology and linguistics."
	},
	"TABREF5": {
	"html": null,
	"type_str": "table",
	"content": "<table/>",
	"num": null,
	"text": "Piotrovsky, who thinks that for the time being thosetwho develop MT systems must not strive to achieve 100% efficiency of text processing and these MT systems must be based not on deductive generative grammar but on deductive-inductive linguistics of text. According to the linguistics of text, MT systems should be developed consecuti.vely from simple algorithms operating with units of the sbrface levels of the text system to more complicated algorithms oriented to the deep levels of the text"
	},
	"TABREF6": {
	"html": null,
	"type_str": "table",
	"content": "<table><tr><td>Piotrovskyaye the principalJinguistic algorithm. All programs of auto-matic processing of linguistic information are based on these c h a r t s . systems was presented by Z. M. Shaljapins. in \"The ARAT system: dictionary, grammar, and their use in automatic analysis\" (Moscow,</td></tr></table>",
	"num": null,
	"text": "A number of reports were read by members of the Speech StatisticsMTt Moscow INTERNATIONAL SEMINAR group (tho work is under the scientific guidance of Dr. R . G. ). Scientists of this group compile dictionaries f o r MT within t h e framework of one scheme i n accordance with the postulates of the linguistics of text; their MT dictionaries consist of two parts: a dictlonary of commonly used words and changeable terminological dictionaries for various fields of science and technology. The input entries of these dictionaries are either single word forms or combinations of several word foms. Morphological and syntactic information is written s t r a i g h t in the dictioqary entries. The volume of syntactic information is relatively small. This information includes indices of grammatical word-class, transitivity-intransitivity for verbs, type of government and some other syntactic characteristics. Dictionaries of input and output languages have charts for coding information and correspondence charts which in thi's MT system The authors of these reports pay great attention to methods of coding information and compressing codes in the computer. The computer realization of such dictionaries makes it possible to get an interlinear version which is not actually a translation but gives t h e consumer the main contents of the original text. During the discussion, many participants of the seminar, criticizing this approach to MT, pointed out that automatic dictionarieg should not be compiled separately but as component parts of MT systems. The coding of lexical information ia an MT system should be preceded by the elaboration of blocks of grammatical analysis and synthesis which use l e x i c a l informatibra and in their turn may affect its composition and coding. h difgerent approach to principles for development of MT"
	},
	"TABREF7": {
	"html": null,
	"type_str": "table",
	"content": "<table><tr><td>). I n t h i s M T</td></tr><tr><td>system t h e disambiguation of homonymy i s n o t singled out i n t o a</td></tr></table>",
	"num": null,
	"text": "block but is carried out during the analysis simultaneously with the fulfillment of other tasks~. Prepositions in the ARAT system are tteated like any other lexical unit of the text. Meaningless preposici~ns, 1 . e . prepositions serving as surface-syntactic indications of strong government and having no influence o* the meaning of the text are eliminated during the transiti.611 \u00a3&om combined syntactic structure (CSC) to semantic representation. Meaningful prepositions, i.e. prepositions having theik own meaning, are elements of the semantic structure and are prederved during the transition from CSC to semantic representation. E. E. Lovtsky (ACT, Moscow) proposed formal means for the descrfption of the syntax of riatural languages. The description is qn oriented graph, in the nodes of which are symbols of syntactic classes and of subgrap The in\u00a3 ormation on the dependency structure of string.: is recorded on the edges of the graph. The description of a natural language syntax, madewith the help of the suggested formal means, can be used in a system of automatic syntactic znalysis. For the analysis of a phrase, it is necessary to find in the graph all occurrences of astringof syntactic classes corresponding to this phrase and to read from the edges of the graph me information about the structure of the string. This procedure is done automatically by a special algorithm, As a result of the analysis we get immediate constituent trees and dependency trees of the analyzed phrase."
	}
	}
	}
	}