| { |
| "paper_id": "A83-1002", |
| "header": { |
| "generated_with": "S2ORC 1.0.0", |
| "date_generated": "2023-01-19T02:11:38.665733Z" |
| }, |
| "title": "PROBLEMS IN NATURAL-LANGUAGE INTERFACE TO DSMS WITH EXAMPLES FROM EUFID", |
| "authors": [ |
| { |
| "first": "Marjorie", |
| "middle": [], |
| "last": "Templeton", |
| "suffix": "", |
| "affiliation": { |
| "laboratory": "", |
| "institution": "System Development Corporation Santa Mortice", |
| "location": { |
| "country": "California" |
| } |
| }, |
| "email": "" |
| }, |
| { |
| "first": "John", |
| "middle": [], |
| "last": "Burger", |
| "suffix": "", |
| "affiliation": { |
| "laboratory": "", |
| "institution": "System Development Corporation Santa Mortice", |
| "location": { |
| "country": "California" |
| } |
| }, |
| "email": "" |
| } |
| ], |
| "year": "", |
| "venue": null, |
| "identifiers": {}, |
| "abstract": "", |
| "pdf_parse": { |
| "paper_id": "A83-1002", |
| "_pdf_hash": "", |
| "abstract": [], |
| "body_text": [ |
| { |
| "text": "For five years the End-User Friendly Interface to Data management (EUFID) project team at System Development Corporation worked on the design and implementation of a Natural-Language Interface (NLI) system that was to be independent of both the application and the database management system. In this paper we describe application, natural-language and database management problems involved in NLI development, with specific reference to the EUFID system as an example. Language\", \"World", |
| "cite_spans": [ |
| { |
| "start": 193, |
| "end": 198, |
| "text": "(NLI)", |
| "ref_id": null |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| }, |
| { |
| "text": "Language\", and \"Data Base Language\" and appear to correspond roughly to the \"external\", \"conceptual\", and \"internal\" views of data as described by C. J. Date", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Model", |
| "sec_num": null |
| }, |
| { |
| "text": "[DATE77]. PHLIQAI can interface to a variety of database structures and DBMSs.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Model", |
| "sec_num": null |
| }, |
| { |
| "text": "The -to be application independent. This means that the program must be table driven.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "5.", |
| "sec_num": null |
| }, |
| { |
| "text": "The tables contain the dictionary and semantic information and are loaded with application-speciflc data.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "5.", |
| "sec_num": null |
| }, |
| { |
| "text": "It was desired that the tables could be constructed by someone other than the EUFID staff, so This language represents, in many ways, the union of the capabilities of many \"target\" DBMS query languages.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "5.", |
| "sec_num": null |
| }, |
| { |
| "text": "that", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "5.", |
| "sec_num": null |
| }, |
| { |
| "text": "The EUFID system consists of three major modules, not counting the DBM3 (see Figure I ).", |
| "cite_spans": [], |
| "ref_spans": [ |
| { |
| "start": 77, |
| "end": 85, |
| "text": "Figure I", |
| "ref_id": null |
| } |
| ], |
| "eq_spans": [], |
| "section": "5.", |
| "sec_num": null |
| }, |
| { |
| "text": "The values which can be be used to convert from one unit of measure to another (e.g., feet to meters).", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "5.", |
| "sec_num": null |
| }, |
| { |
| "text": "These data are used by the run-time modules which map and translate the tree-structured output of the analyzer to IL on the actual group/field names of the database, and then co the language of the DBMS.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "5.", |
| "sec_num": null |
| }, |
| { |
| "text": "These modules are discussed in the next sections.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "5.", |
| "sec_num": null |
| }, |
| { |
| "text": "The EUFID Analyzer \"root\" to the database structure.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "2.", |
| "sec_num": null |
| }, |
| { |
| "text": "Access may be made from any relation to any other relation as long as there is a field in each of the two relations which has the same \"domain\" (set of values). These are discussed below.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "2.", |
| "sec_num": null |
| }, |
| { |
| "text": "REPorting", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "AIREP (ADP Incident", |
| "sec_num": null |
| }, |
| { |
| "text": "An English word may have more than one definition without complicating the analysis strategy. For example, \"ship\" as a vessel and as a verb meaning \"to send\" can be defined in the same dictionary.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "AIREP (ADP Incident", |
| "sec_num": null |
| }, |
| { |
| "text": "Words used as database values, such as names, may also have multiple definitions, e.g., \"New York\" used as the name of both a city and a state. of which must ~oin both the company (\"c =) and the warehouse ('w') relations to the =cw\" relation. Values can also be used in a query to qualify or select certain records for output, e.g., in the above question \"North Hills\" and \"Superior\" are values that must be represented in the query to the DBMS.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "AIREP (ADP Incident", |
| "sec_num": null |
| }, |
| { |
| "text": "As long as the alphanumeric values used in a particular database field are the same as words in the English questions, there are no difficult problems involved in recognizing values as selectors in a query.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "AIREP (ADP Incident", |
| "sec_num": null |
| }, |
| { |
| "text": "There are three basic ways to recognize these value words in a question. They can be explicitly listed in the dictionary, recognized by a pattern or context, or found in the database itself.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "AIREP (ADP Incident", |
| "sec_num": null |
| }, |
| { |
| "text": "If the value words are stored in the dictionary, they can be subject to spelling correction because the spelling corrector uses the dictionary to locate words which are a close match to unrecognized words in a question. In APPLICANT, however, each applicant has a set of \"specialties\" such as \"computer programmer\", \"accounting clerk\", or \"gardener\". These are all stored as values of the specialty field in the database. Essentially syntactic information is used only when needed to resolve ambiguity.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "AIREP (ADP Incident", |
| "sec_num": null |
| }, |
| { |
| "text": "The language features that this technique has to handle are common to any NLI, and some of the problem areas are described in the following sections.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "AIREP (ADP Incident", |
| "sec_num": null |
| }, |
| { |
| "text": "To support natural interaction it is desirable to allow the use of anaphoric reference and elliptical constructions across sentence sequences, such as \"What applicants know Fortran and C?\", \"Which of them live in California?\", \"In Nevada?\", \"How many know Pascal?'. One of the biggest problems is to define the scope of the reference in such cases.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "I. Anaphora and Ellipsis", |
| "sec_num": null |
| }, |
| { |
| "text": "In the example, it is not clear whether the user wishes to retrieve the set of all applicants who know Pascal or only the subset who live in Nevada.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "I. Anaphora and Ellipsis", |
| "sec_num": null |
| }, |
| { |
| "text": "One solution is to provide commands that allow users to define subsets of the database to which to address questions. This removes the ambiguity and speeds up retrieval time on a large database.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "I. Anaphora and Ellipsis", |
| "sec_num": null |
| }, |
| { |
| "text": "However, it moves the NLI interaction toward that of a structured query language, and forces the user to be aware of the level of subset being accessed. ", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "I. Anaphora and Ellipsis", |
| "sec_num": null |
| }, |
| { |
| "text": "In normal NLI interaction users may wish to ask \"yes/no\" questions, yet no DBMS has the ability to answer \"yes\" or \"no\" explicitly.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Yes/No Questions", |
| "sec_num": "3." |
| }, |
| { |
| "text": "The EUFID mapper maps a yes/no question into a query which will retrieve some data, such as an \"output identifier\" or default name for a concept, if the answer is \"yes\" and no data if the answer if \"no\".", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Yes/No Questions", |
| "sec_num": "3." |
| }, |
| { |
| "text": "However, the answer may be \"no\" for several reasons. ", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Yes/No Questions", |
| "sec_num": "3." |
| }, |
| { |
| "text": "The scope of conjunctions is a difficult problem for any parsing or analyzing algorithm.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Conjunctions", |
| "sec_num": "4." |
| }, |
| { |
| "text": "The natural-language use of \"and\" and \"or\" does not necessarily correspond to the logical meaning, as in the question \"List the applicants who live in California and Arizona.\".", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Conjunctions", |
| "sec_num": "4." |
| }, |
| { |
| "text": "Multiple conjunctions in a single question can be ambiguous as in \"which minority and female applicants know Fortran and Cobol?'. This could be interpreted with logical \"and\" or with logical \"or\" as in \"Which applicants who are minority or female know either Fortran or Cobol?\".", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Conjunctions", |
| "sec_num": "4." |
| }, |
| { |
| "text": "The EUFID mapper will change English \"and\" to logical \"or\" when the two phrases within the scope of the conjunction are values for the same field. In the example above, an applicant has only one state of residence. uncertain whether they should be returned in the answer. It is also difficult to take a complement of a set of data using the many data management systems that do not support set operators between relations.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Conjunctions", |
| "sec_num": "4." |
| }, |
| { |
| "text": "Questions which require a \"yes\" or \"no\" response are difficult to answer because often the \"no\" is due to a presupposition which is invalid. This is especially true with negation.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Conjunctions", |
| "sec_num": "4." |
| }, |
| { |
| "text": "For example, if the user asks, \"Does every company in North Hills except Supreme use NH2?\", the answer may be \"no\" because Supreme is not in North Hills.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Conjunctions", |
| "sec_num": "4." |
| }, |
| { |
| "text": "The current implementation of EUFID does not allow explicit negation, although some negative concepts are handled such as \"What companies ship to companies other than Colonial?\". \"Other than\" is interpreted as the \"!-\" operator in exactly the same way that \"greater than\" is interpreted as \">\".", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Conjunctions", |
| "sec_num": "4." |
| }, |
| { |
| "text": "Many questions make perfect sense semantically but are difficult to map into DBMS queries because of the database structure.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "C. INTERPRETATION AND DATABASE ISSUES", |
| "sec_num": null |
| }, |
| { |
| "text": "The problems become worse when access is through an NLI because of increased expectations on the part of the user and because it may be difficult for a help system adequately to describe the problem to the user who is unaware of the database structure.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "C. INTERPRETATION AND DATABASE ISSUES", |
| "sec_num": null |
| }, |
| { |
| "text": "Negative requests may contain explicit negative words such as \"not\" and \"never\" or may contain implicit negatives such as \"only\", \"except\" and \"other than\" [OLNE78] . In the open world database, which we encounter most of the time, a response of \"not that this database knows of\" might be more appropriate.", |
| "cite_spans": [ |
| { |
| "start": 156, |
| "end": 164, |
| "text": "[OLNE78]", |
| "ref_id": null |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Nepption", |
| "sec_num": null |
| }, |
| { |
| "text": "The design of the IL is critical. It must be rich enough to support retrieval from all the underlying DBMSs. However, if it contains capabilities that do not exist in a specific DBMS, it is difficult to describe this deficiency to the user.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "I. IL Limitations", |
| "sec_num": null |
| }, |
| { |
| "text": "In APPLICANT, the user cannot get both the major and minor fields of study by asking \"List applicants and field of study\", because a limitation in the EUFID IL prevents making two joins between education and subject records.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "I. IL Limitations", |
| "sec_num": null |
| }, |
| { |
| "text": "This problem was corrected in a subsequent version of IL with the addition of a \"range\" state- in the proper case, the value will not match.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "I. IL Limitations", |
| "sec_num": null |
| }, |
| { |
| "text": "The NLI user is not expected to understand exactly how data is stored, and yet must understand something about the granularity of the data.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Granularit~ Differences", |
| "sec_num": "3." |
| }, |
| { |
| "text": "Time fields often cause problems because time may be given by year or by fractions of a second.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Granularit~ Differences", |
| "sec_num": "3." |
| }, |
| { |
| "text": "Users may make time comparisons that require more granularity than is stored in the database.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Granularit~ Differences", |
| "sec_num": "3." |
| }, |
| { |
| "text": "For example, the user can ask \"What incidents were reported at SAC while system release 3.4 was installed?\". If incidents were reported by day but system release dates were given by month, the system would return incidents which occurred in the days of the month before the system release was installed.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Granularit~ Differences", |
| "sec_num": "3." |
| }, |
| { |
| "text": "A very simple question in English can turn into a very complicated request in the query language if it involves retrieval of data which must be used for qualification in another part of the same query.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Nested Queries", |
| "sec_num": "4." |
| }, |
| { |
| "text": "In IL these are called \"nested queries\".", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Nested Queries", |
| "sec_num": "4." |
| }, |
| { |
| "text": "Most often some qualification needs to be done both \"inside\" and \"outside\" the clause of the query that does the internal retrieve. There are problems that need to be solved on both the front end, the parsing of the English question, and the back end, the translation of the question into a data management system query.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Nested Queries", |
| "sec_num": "4." |
| }, |
| { |
| "text": "It is important to understand the types of requests, types of functions, and types of databases that can be supported by a specific NLI. ", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Nested Queries", |
| "sec_num": "4." |
| }, |
| { |
| "text": "the users of the NLI should have a common use for the data and a common vlew of the data, and 7.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "6.", |
| "sec_num": null |
| }, |
| { |
| "text": "there must be some user who understands the questions that will be asked and is available to work with the developers of the NLI.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "6.", |
| "sec_num": null |
| }, |
| { |
| "text": "We believe that current system development is limited by the need for good semantic modelling techniques and the length of time needed to build the knowledge base required to interface with a new application. When the knowledge base for the NLI is developed, the database as well as sample input must be considered in the design.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "6.", |
| "sec_num": null |
| }, |
| { |
| "text": "Parsing of questions to a database cannot be divorced from the database contents since semantic interpretation can only be determined in the context of that database. On the other hand, a robust system cannot be developed by considering only database structure and content, because the range of the questions allowed would not accurately reflect the user view of the application and also would not account for all the information that is inferred at some level.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "6.", |
| "sec_num": null |
| } |
| ], |
| "back_matter": [ |
| { |
| "text": "We would like to acknowledge", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "ACKNOWLEDGEMENTS", |
| "sec_num": null |
| } |
| ], |
| "bib_entries": { |
| "BIBREF0": { |
| "ref_id": "b0", |
| "title": "The Theory of Parsing, Translation, and Compiling", |
| "authors": [ |
| { |
| "first": "A", |
| "middle": [ |
| "V" |
| ], |
| "last": "Aho", |
| "suffix": "" |
| }, |
| { |
| "first": "J", |
| "middle": [ |
| "D" |
| ], |
| "last": "Ullman", |
| "suffix": "" |
| } |
| ], |
| "year": 1972, |
| "venue": "", |
| "volume": "I: Parsing", |
| "issue": "", |
| "pages": "314--337", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Aho, A. V. and J. D. Ullman, \"The Theory of Parsing, Translation, and Compiling\", Vol. I: Parsing, Prentice-Hall, 1972, pp. 314-23Z.", |
| "links": null |
| }, |
| "BIBREF1": { |
| "ref_id": "b1", |
| "title": "Semantic Database Mapping in EUFID", |
| "authors": [ |
| { |
| "first": "J", |
| "middle": [ |
| "F" |
| ], |
| "last": "Burger", |
| "suffix": "" |
| } |
| ], |
| "year": null, |
| "venue": "Proceedings of the 198Z ACM/SIGMOD Conference, ~3\"n~-a'-o~caY-'-Ca~", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Burger, J. F., \"Semantic Data- base Mapping in EUFID\", Proceedings of the 198Z ACM/SIGMOD Conference, ~3\"n~-a'-o~caY-'-Ca~., May 14-16, 198ff.", |
| "links": null |
| } |
| }, |
| "ref_entries": { |
| "TABREF6": { |
| "type_str": "table", |
| "text": "The mapper, despite its many limitations, can correctly map almost all trees output by the analyzer.", |
| "num": null, |
| "html": null, |
| "content": "<table><tr><td/><td>The</td><td colspan=\"2\">mapper</td><td colspan=\"2\">can</td><td colspan=\"2\">translate</td><td>\"user</td></tr><tr><td colspan=\"2\">values\"</td><td colspan=\"2\">(e.g.,</td><td colspan=\"3\">\"Russian\")</td><td>to database</td></tr><tr><td colspan=\"6\">values (e.g., \"USSR\"),</td><td>and</td><td>convert</td><td>one</td></tr><tr><td>unit</td><td colspan=\"7\">of measure (e.g., feet) to another</td></tr><tr><td colspan=\"4\">(e.g., meters).</td><td/><td/><td/></tr><tr><td/><td colspan=\"5\">EUFID can interface</td><td colspan=\"2\">to very</td><td>complex</td></tr><tr><td colspan=\"3\">relational</td><td>and</td><td colspan=\"4\">CODASYL-type</td><td>databases</td></tr><tr><td colspan=\"7\">having difficult navigation</td><td>and</td><td>parallel</td></tr><tr><td colspan=\"3\">structures.</td><td>In</td><td colspan=\"4\">the AIREP application a</td></tr><tr><td colspan=\"2\">consistent</td><td/><td colspan=\"5\">WWDMS navigational methodology</td></tr><tr><td>is</td><td>used</td><td colspan=\"6\">to access non-key records.</td><td>The</td></tr><tr><td colspan=\"8\">system can also map to the parallel,</td><td>but</td></tr><tr><td>not</td><td colspan=\"3\">identical,</td><td colspan=\"4\">structures for duplicate</td></tr><tr><td colspan=\"7\">and historical incidents.</td></tr><tr><td/><td colspan=\"7\">In the INGRES applications, EUFID is</td></tr><tr><td>able</td><td>to</td><td colspan=\"6\">use and correctly map to =rela-</td></tr><tr><td colspan=\"2\">tionship</td><td colspan=\"3\">relations\"</td><td colspan=\"3\">which relate</td><td>two</td><td>or</td></tr><tr><td>more</td><td colspan=\"2\">other</td><td colspan=\"3\">relations.</td><td colspan=\"2\">For example, the</td></tr><tr><td colspan=\"5\">METRO relation =cw\"</td><td colspan=\"3\">contains</td><td>a</td><td>company</td></tr><tr><td colspan=\"8\">name, a warehouse name, and a date.</td><td>This</td></tr><tr><td colspan=\"2\">represents</td><td/><td colspan=\"2\">the initial</td><td/><td colspan=\"2\">business</td><td>contact.</td></tr><tr><td colspan=\"2\">A user</td><td colspan=\"2\">might</td><td>ask,</td><td/><td colspan=\"2\">=When did Colonial</td></tr><tr><td>start</td><td colspan=\"6\">to do business with</td><td>Superior? = or</td></tr><tr><td colspan=\"8\">\u2022 When did business begin between Colonial</td></tr><tr><td colspan=\"6\">and Superior? = , either</td><td/></tr><tr><td/><td/><td/><td/><td/><td/><td/><td>It</td><td>is</td><td>able</td><td>to</td></tr><tr><td/><td/><td/><td/><td/><td/><td/><td>handle English conjunctions, mapping them</td></tr><tr><td/><td/><td/><td/><td/><td/><td/><td>appropriately to logical ANDs or ORs, and</td></tr><tr><td/><td/><td/><td/><td/><td/><td/><td>understanding</td><td>that</td><td>some \"ands\" may need</td></tr><tr><td/><td/><td/><td/><td/><td/><td/><td>to be interpreted as</td><td>OR</td><td>and</td><td>vice-versa</td></tr><tr><td/><td/><td/><td/><td/><td/><td/><td>under</td><td>certain circumstances.</td><td>It is able</td></tr><tr><td/><td/><td/><td/><td/><td/><td/><td>to generate calls</td><td>on</td><td>DBMS</td><td>calculations</td></tr><tr><td/><td/><td/><td/><td/><td/><td/><td>(e.g.,</td><td>average)</td><td>and</td><td>user-defined func-</td></tr><tr><td/><td/><td/><td/><td/><td/><td/><td>tions</td><td>(e.g.,</td><td>marine</td><td>great-circle</td><td>dis-</td></tr><tr><td/><td/><td/><td/><td/><td/><td/><td>tance) if the user-function exists and is</td></tr><tr><td/><td/><td/><td/><td/><td/><td/><td>supported by the DBMS.</td></tr><tr><td/><td/><td/><td/><td/><td/><td/><td>Questions involving time are</td><td>inter-</td></tr><tr><td/><td/><td/><td/><td/><td/><td/><td>preted</td><td>in</td><td>a</td><td>reasonable way.</td><td>Functions</td></tr><tr><td/><td/><td/><td/><td/><td/><td/><td>are defined for \"between\" and \"during\" in</td></tr><tr><td/><td/><td/><td/><td/><td/><td/><td>the</td><td>METRO application.</td><td>The AIREP appli-</td></tr><tr><td/><td/><td/><td/><td/><td/><td/><td>cation allows time</td><td>comparisons</td><td>such</td><td>as</td></tr><tr><td/><td/><td/><td/><td/><td/><td/><td>\"What</td><td>system</td><td>was</td><td>running when incident</td></tr><tr><td/><td/><td/><td/><td/><td/><td/><td>J123 occurred\" which require</td><td>a</td><td>test</td><td>to</td></tr><tr><td/><td/><td/><td/><td/><td/><td/><td>see</td><td>if</td><td>a</td><td>point</td><td>in</td><td>time</td><td>is within an</td></tr><tr><td/><td/><td/><td/><td/><td/><td/><td>interval.</td></tr></table>" |
| }, |
| "TABREF14": { |
| "type_str": "table", |
| "text": "language access to databases. It is not clear that such a system exists for general use [0SI79].", |
| "num": null, |
| "html": null, |
| "content": "<table><tr><td/><td>generate</td><td>the IL query</td></tr><tr><td>D. such required OVERALL NLI DESIGN There are several as navigational in a naval database. affect 4. Support for Metadata the selection of applications for problems calculations that the NLI. Some databases and data manage-ment Metadata is data about the data in systems may not be appropriate tar-gets for natural-language the database. It would be able to tell interfaces. Some DBMS the user of the METRO application, for functions may be difficult to support. example, the kind of information the It is important to have a clear understanding of database has for warehouses and other these problems so that the NLI can mediate between the entities in the application. Such meta-user view, as represented by the data might be extensions of active natural-language questions, and the integrated data dictionaries now avail-underlying database structure. able in some DBMSs.</td><td colspan=\"3\">retrieve where (cct.date [cct.scname] \u2022 198~) and (cct.lf >-{retrieve [avg (cct.lf)] where (cct.date -1980)}) Here, =cct\" is the name of the company-to-company transaction relation. \"Scname\" is the name of a shipping com-pany in this relation. Note again that the qualification on \"198~\" needs to be done both inside and outside the nested part of the query. In the query language for INGRES map queries and to explain problems the user when the mapping cannot be made. to However, there can be \"reasonable\" queries that cannot be answered directly because of the database structure. Hierarchical DBMSs present the most prob-lems with navigation because access must start from the root. For example, if the APPLICANT database were under an hierarchical DBMS, the question \"List the specialties for each applicant\" could be answered directly but not \"What are the specialties?\" as there would be no way to particular applicant records. get to the specialty records except via</td></tr><tr><td>In an application-level i. ~ Design Consideration system user should be able to query the metadata the</td><td colspan=\"2\">such very similtar a request to the IL expressions. is expressed in a manner For An array allows more than</td><td>one</td></tr><tr><td>For any database there are to learn about the structure of the data-natural-language base. A different mode, such as the questions that cannot be inter-preted because the concepts involved menus used by the EUFID help system, lle outside the world of the database. could be used to access metadata, or Ques-tions can also involve structural English language questions to both meta com-plexity information and the database could be that is not representable in the DBMS query language. supported. A particularly dif-ficult decision in the overall design of 5. Updates an NLI is the issue of where in the chain of events of processing a user's question Some potential users would like a into a DBMS query to trap these questions natural-language interface to include the and stop processing. capability to update the database.</td><td colspan=\"3\">WWDMS a erated. answer result for usa in very complex procedure In all cases, the DBMS needs is gen-to the inner request and save the qualifying the outer request. There are many database manage-ment systems that cannot handle such questions and these IL statements cannot be translated into the system's query language. 5. Inconsistency In Retrieval The NLI presents a uniform view of instance of a field or set of fields in a single record. There may be arrays of values or even arrays of sets of values in nonrelatlonal databases. When the user retrieves a field that is an array the DBMS requires a subscript into the array. Either the user must specifly this subscript or the NLI must map to all members of the array with a test for missing data. 3. Class of DBMS to Supp%rt</td></tr><tr><td>One approach is to decide that if Currently, updating through any high question level view of the database should be is not meaningful to the world a avoided, especially when the view con-of the database it should not be meaning-ful to the NLI and, therefore, tains joins or derlve4 data, because of not analyzable on semantic grounds. risk of inadvertently entering the Another assumes that if the NLI can analyze a incorrectly-interpreted data.</td><td colspan=\"3\">all databases cult to truly mask all differences in the and DBMSs, but it is diffi-behavior of the DBMSS because they do not all process the equivalent query in the same way. For example, when data are retrieved from two relations in a rela-tional database, the two relations must For systems such as EUFID, the data-base must be organized within a data managdment system so that the data is structured and individual fields are named. If the data is just text, the EUFID approach cannot be used.</td></tr><tr><td>For example, the at SAC had the downtime?\" from our AIREP appli-\"What incident is expressed in IL as retrieve [INCA. ID] question longest cation where (INCA.SITENAME = \"SAC\") and (INCA.DNTM -Codd made good use of the procedures of the RENDEZVOUS system to avoid questions that the could not handle, as generation of DBMS queries that did represent the user's intent. tem, however, requires a very semantic base the database) in order to make meaningful (much larger than that of large Such a sys-not well as avoiding DBMS [CODD74] dialogue question that cannot be asked of the database, V SUMMARY AND CONCLUSIONS it has a much better chance of describing to the user what is wrong with the question and how it might For many years, researchers have be rephrased to get the desired information. been attempting to build robust systems for natural-</td><td colspan=\"3\">be Joined on answer forms a new relation which may be a common attribute. The displayed to the user or stored. Since the join clause acts as qualification, a record (tuple) in either relation which has no corresponding tuple relation does not participate in the result. This is a different concept from the hierarchical and network models where the system retrieves all records from a master record and then retrieves corresponding records from a subfile. This difference can cause anomalies with retrieval. For example, in a pure rela-tional system \"List applicants and thei~ the returned answer. database for parsing or for processing query programs without any access to the must have the capability to generate end is to be added to such a DBMS, i~ return for many minutes. If an Nil front batch job queue and answers may not language, but queries are entered into a tive. WWDMS [HONE76] has a user query in the other Current NLI systems are designed to be used interactively by a user, which means that the DBMS should also have an interactive query language. However, noc all data management systems are interac-</td></tr><tr><td>[retrieve [ max (INCA.DNTM)] where (INCA.SITENAME = \"SAC\")}) The nested part of the query is enclosed in braces. \"INCA\" is the database name of the active incident records. Notice that removing the \"INCA.SITENAME = 'SAC'\" clause from either the inner or outer query would result in an incorrect formu-lation of the question. A similar example from the communication with the user during the dialogue. 2. Class of Database to Support Some databases are simply not good candidates for an NLI because of charac-teristics mentioned in previous sections such as many retrieve-only fields, or domains that have a high update rate but cannot be recognized by a pattern. METRO application is the question, \"What com-There are also some structural prob-</td><td colspan=\"3\">interviews\" applicants together with would who their be have The query language treated had should as \"List interviews support operations equivalent to the relational interview informa-tion.\" A hierarchical or network DBMS would treat operations of select, project, and join. Also, the query language should support it as \"List all applicants (whether or not they have been inter-viewed) plus any interview information that exists.\" This second interpreta-tion is more likely to be the correct one. some arithmetic capability. Most have aggregate functions such as SUM and COUNT. WWDMS does not have an easy-to-use average operation, but it does have a procedural language with arithmetic operators so that EUFID can produce a \"query\" that procedurally calculates an average.</td></tr><tr><td>pany shipped more than the average amount of light freight in 198~?\" which will lems chat must be recognized. If the database contains \"flat\" files about one basic entity, it is reasonably easy to 13</td><td colspan=\"3\">Basic calculations ported such as \"age=today-birthdate\". should be is also desirable to be able to call spe-sup-It cial functions to do complex calculations</td></tr></table>" |
| } |
| } |
| } |
| } |
