Hugging Face's logo

RaguTeam
/
BertRE

Text Classification
Transformers
Safetensors
Russian
bert
relation-extraction
ru-bert
nerel
text-embeddings-inference
Model card Files
xet
Community
IlyaMyzk commited on
Commit
dca79cc
·
verified ·
1 Parent(s): 9989b87

Upload 9 files

Browse files
Files changed (9) hide show
  1. README.md +94 -0
  2. added_tokens.json +6 -0
  3. config.json +140 -0
  4. model.safetensors +3 -0
  5. relation_constraints.json +1244 -0
  6. special_tokens_map.json +67 -0
  7. tokenizer.json +0 -0
  8. tokenizer_config.json +96 -0
  9. vocab.txt +0 -0
README.md ADDED
@@ -0,0 +1,94 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ language:
4
+ - ru
5
+ library_name: transformers
6
+ pipeline_tag: text-classification
7
+ tags:
8
+ - relation-extraction
9
+ - ru-bert
10
+ - nerel
11
+ datasets:
12
+ - nerel
13
+ ---
14
+
15
+ # Russian Relation Extraction Model
16
+
17
+ This model is trained for the task of Relation Extraction between named entities in Russian text. It takes a piece of text and two marked entities within it as input and predicts the most likely semantic relationship between them (e.g., `WORKS_AS`, `WORKPLACE`, `SPOUSE`, etc.).
18
+
19
+ The model is based on the R-BERT architecture and has been fine-tuned on the [Nerel](https://github.com/nerel-ds/nerel) dataset.
20
+
21
+ ## Model Details
22
+
23
+ * **Base Model:** `DeepPavlov/rubert-base-cased`
24
+ * **Architecture:** R-BERT. The model leverages not only the `[CLS]` token representation but also the averaged representations of each entity's tokens, along with embeddings for their types (e.g., `PERSON`, `ORGANIZATION`). This allows the model to better understand the context and the nature of the interacting entities.
25
+ * **Language:** Russian
26
+
27
+ ## How to Use
28
+
29
+ This model is intended to be used in a pipeline, following a Named Entity Recognition (NER) model. After a NER model has identified entities in the text, this model can be used to predict relationships between all possible pairs of those entities.
30
+
31
+ For practical use, the easiest way to deploy this model is via the provided Docker container, which exposes a REST API.
32
+
33
+ ### Deployment with Docker
34
+
35
+ 1. **Pull the Docker image:**
36
+ ```bash
37
+ docker pull mrpzzios/bertre:1.3
38
+ ```
39
+ 2. **Run the container (with GPU acceleration):**
40
+ ```bash
41
+ docker run -d -p 8000:8000 --name bertre-api mrpzzios/bertre:1.3
42
+ ```
43
+ 3. **Send a request to the API:**
44
+ ```bash
45
+ curl -X POST "http://localhost:8000/predict" \
46
+ -H "Content-Type: application/json" \
47
+ -d '{
48
+ "chunks": ["Президент Башкирии Муртаза Рахимов решил поменять главу своей администрации."],
49
+ "entities_list": [
50
+ [[19, 34, "PERSON"], [0, 18, "PROFESSION"], [50, 75, "PROFESSION"]]
51
+ ]
52
+ }'
53
+ ```
54
+
55
+ ## Training and Inference Methodology
56
+
57
+ ### Training Process (Multi-Label Approach)
58
+
59
+ The model was trained on the Nerel dataset using a multi-label formulation, which is crucial for handling cases where a single pair of entities can have multiple valid relationships.
60
+
61
+ * **Label Representation:** The relationship labels for each training example were converted into a **binary vector (bitmask)**. In this vector, each index corresponds to a specific relation type. A value of `1` indicates that the relation exists, while `0` indicates it does not.
62
+
63
+ * **Loss Function:** Consequently, **`BCEWithLogitsLoss`** was used as the loss function. This function is ideal for multi-label tasks as it evaluates each output logit from the model independently against the corresponding value in the target bitmask. This teaches the model to assess the "evidence" for each relationship type on its own merits, rather than forcing it to choose just one during training. This results in a more nuanced understanding of the data.
64
+
65
+ ### Inference Process (Single-Label Output)
66
+
67
+ During inference, the model produces a vector of logits (raw scores) for all possible relationship types. To provide a single, most confident prediction, the following steps are taken:
68
+
69
+ 1. The model identifies the **single relationship with the highest logit score**.
70
+ 2. The confidence score for this winning relationship (the `relation_strength`) is calculated by applying a **Sigmoid function** to its logit value. This converts the raw score into a more interpretable value between 0 and 1.
71
+
72
+ This approach combines the robust learning of a multi-label setup with a decisive single-label output, making it practical for downstream applications that expect one definitive relationship per entity pair.
73
+
74
+ ## Constrained Decoding
75
+
76
+ During inference and evaluation, a schema of type constraints derived from the Nerel dataset's annotation guidelines was applied. This prevents the model from predicting logically impossible relations (e.g., a `SPOUSE` relation between a `PERSON` and an `ORGANIZATION`), which significantly improves prediction precision.
77
+
78
+ ## Evaluation
79
+
80
+ The model was evaluated on the validation split of the Nerel dataset. The following metrics were achieved (macro average):
81
+
82
+ | Metric | Value |
83
+ |---------------|---------|
84
+ | **F1-score** | 0.7500 |
85
+ | **Precision** | 0.8286 |
86
+ | **Recall** | 0.7246 |
87
+
88
+ *Note: These metrics were obtained with constrained decoding applied during evaluation.*
89
+
90
+ ## Limitations and Bias
91
+
92
+ * The model's performance is highly dependent on the quality of the upstream Named Entity Recognition (NER) model. Errors from the NER stage will propagate and cause errors in relation extraction.
93
+ * Performance may degrade on texts from domains that differ significantly from the news and encyclopedic articles found in the Nerel dataset.
94
+ * Like many language models, this model may reproduce statistical biases present in its training data. For example, it might associate certain professions more strongly with a particular gender.
added_tokens.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "</e1>": 119548,
3
+ "</e2>": 119550,
4
+ "<e1>": 119547,
5
+ "<e2>": 119549
6
+ }
config.json ADDED
@@ -0,0 +1,140 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "./model",
3
+ "architectures": [
4
+ "BertRE"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "classifier_dropout": null,
8
+ "directionality": "bidi",
9
+ "dropout_rate": 0.0,
10
+ "dtype": "float32",
11
+ "entity_type_embedding_dim": 128,
12
+ "hidden_act": "gelu",
13
+ "hidden_dropout_prob": 0.1,
14
+ "hidden_size": 768,
15
+ "id2label": {
16
+ "0": "LABEL_0",
17
+ "1": "LABEL_1",
18
+ "2": "LABEL_2",
19
+ "3": "LABEL_3",
20
+ "4": "LABEL_4",
21
+ "5": "LABEL_5",
22
+ "6": "LABEL_6",
23
+ "7": "LABEL_7",
24
+ "8": "LABEL_8",
25
+ "9": "LABEL_9",
26
+ "10": "LABEL_10",
27
+ "11": "LABEL_11",
28
+ "12": "LABEL_12",
29
+ "13": "LABEL_13",
30
+ "14": "LABEL_14",
31
+ "15": "LABEL_15",
32
+ "16": "LABEL_16",
33
+ "17": "LABEL_17",
34
+ "18": "LABEL_18",
35
+ "19": "LABEL_19",
36
+ "20": "LABEL_20",
37
+ "21": "LABEL_21",
38
+ "22": "LABEL_22",
39
+ "23": "LABEL_23",
40
+ "24": "LABEL_24",
41
+ "25": "LABEL_25",
42
+ "26": "LABEL_26",
43
+ "27": "LABEL_27",
44
+ "28": "LABEL_28",
45
+ "29": "LABEL_29",
46
+ "30": "LABEL_30",
47
+ "31": "LABEL_31",
48
+ "32": "LABEL_32",
49
+ "33": "LABEL_33",
50
+ "34": "LABEL_34",
51
+ "35": "LABEL_35",
52
+ "36": "LABEL_36",
53
+ "37": "LABEL_37",
54
+ "38": "LABEL_38",
55
+ "39": "LABEL_39",
56
+ "40": "LABEL_40",
57
+ "41": "LABEL_41",
58
+ "42": "LABEL_42",
59
+ "43": "LABEL_43",
60
+ "44": "LABEL_44",
61
+ "45": "LABEL_45",
62
+ "46": "LABEL_46",
63
+ "47": "LABEL_47",
64
+ "48": "LABEL_48",
65
+ "49": "LABEL_49"
66
+ },
67
+ "initializer_range": 0.02,
68
+ "intermediate_size": 3072,
69
+ "label2id": {
70
+ "LABEL_0": 0,
71
+ "LABEL_1": 1,
72
+ "LABEL_10": 10,
73
+ "LABEL_11": 11,
74
+ "LABEL_12": 12,
75
+ "LABEL_13": 13,
76
+ "LABEL_14": 14,
77
+ "LABEL_15": 15,
78
+ "LABEL_16": 16,
79
+ "LABEL_17": 17,
80
+ "LABEL_18": 18,
81
+ "LABEL_19": 19,
82
+ "LABEL_2": 2,
83
+ "LABEL_20": 20,
84
+ "LABEL_21": 21,
85
+ "LABEL_22": 22,
86
+ "LABEL_23": 23,
87
+ "LABEL_24": 24,
88
+ "LABEL_25": 25,
89
+ "LABEL_26": 26,
90
+ "LABEL_27": 27,
91
+ "LABEL_28": 28,
92
+ "LABEL_29": 29,
93
+ "LABEL_3": 3,
94
+ "LABEL_30": 30,
95
+ "LABEL_31": 31,
96
+ "LABEL_32": 32,
97
+ "LABEL_33": 33,
98
+ "LABEL_34": 34,
99
+ "LABEL_35": 35,
100
+ "LABEL_36": 36,
101
+ "LABEL_37": 37,
102
+ "LABEL_38": 38,
103
+ "LABEL_39": 39,
104
+ "LABEL_4": 4,
105
+ "LABEL_40": 40,
106
+ "LABEL_41": 41,
107
+ "LABEL_42": 42,
108
+ "LABEL_43": 43,
109
+ "LABEL_44": 44,
110
+ "LABEL_45": 45,
111
+ "LABEL_46": 46,
112
+ "LABEL_47": 47,
113
+ "LABEL_48": 48,
114
+ "LABEL_49": 49,
115
+ "LABEL_5": 5,
116
+ "LABEL_6": 6,
117
+ "LABEL_7": 7,
118
+ "LABEL_8": 8,
119
+ "LABEL_9": 9
120
+ },
121
+ "layer_norm_eps": 1e-12,
122
+ "max_position_embeddings": 512,
123
+ "model_type": "bert",
124
+ "num_attention_heads": 12,
125
+ "num_entity_types": 31,
126
+ "num_hidden_layers": 12,
127
+ "output_past": true,
128
+ "pad_token_id": 0,
129
+ "pooler_fc_size": 768,
130
+ "pooler_num_attention_heads": 12,
131
+ "pooler_num_fc_layers": 3,
132
+ "pooler_size_per_head": 128,
133
+ "pooler_type": "first_token_transform",
134
+ "position_embedding_type": "absolute",
135
+ "torch_dtype": "float32",
136
+ "transformers_version": "4.49.0",
137
+ "type_vocab_size": 2,
138
+ "use_cache": true,
139
+ "vocab_size": 119551
140
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8bda6204ad54af0723af6d96cfa3bf282fce85151f1a08078db7402a22856111
3
+ size 719081448
relation_constraints.json ADDED
@@ -0,0 +1,1244 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "<ENTITY>,<ENTITY>": [
3
+ "OWNER_OF",
4
+ "ABBREVIATION",
5
+ "ORIGINS_FROM",
6
+ "RELATIVE",
7
+ "ALTERNATIVE_NAME"
8
+ ],
9
+ "PERSON,<ENTITY>": [
10
+ "PRODUCES",
11
+ "KNOWS"
12
+ ],
13
+ "PROFESSION,<ENTITY>": [
14
+ "PRODUCES",
15
+ "KNOWS"
16
+ ],
17
+ "<ENTITY>,AGE": [
18
+ "AGE_IS"
19
+ ],
20
+ "PERSON,AGE": [
21
+ "AGE_DIED_AT"
22
+ ],
23
+ "PROFESSION,AGE": [
24
+ "AGE_DIED_AT"
25
+ ],
26
+ "PERSON,AWARD": [
27
+ "AWARDED_WITH"
28
+ ],
29
+ "PROFESSION,AWARD": [
30
+ "AWARDED_WITH"
31
+ ],
32
+ "ORGANIZATION,AWARD": [
33
+ "AWARDED_WITH",
34
+ "PART_OF"
35
+ ],
36
+ "WORK_OF_ART,AWARD": [
37
+ "AWARDED_WITH",
38
+ "PART_OF"
39
+ ],
40
+ "NATIONALITY,AWARD": [
41
+ "AWARDED_WITH"
42
+ ],
43
+ "PERSON,CITY": [
44
+ "PLACE_RESIDES_IN",
45
+ "PLACE_OF_BIRTH",
46
+ "PLACE_OF_DEATH",
47
+ "LOCATED_IN",
48
+ "WORKPLACE"
49
+ ],
50
+ "PERSON,COUNTRY": [
51
+ "PLACE_RESIDES_IN",
52
+ "PLACE_OF_BIRTH",
53
+ "LOCATED_IN",
54
+ "WORKPLACE",
55
+ "MEMBER_OF",
56
+ "PLACE_OF_DEATH"
57
+ ],
58
+ "PERSON,DISTRICT": [
59
+ "PLACE_RESIDES_IN",
60
+ "PLACE_OF_BIRTH",
61
+ "PLACE_OF_DEATH",
62
+ "LOCATED_IN",
63
+ "WORKPLACE"
64
+ ],
65
+ "PERSON,FACILITY": [
66
+ "PLACE_RESIDES_IN",
67
+ "PLACE_OF_BIRTH",
68
+ "PLACE_OF_DEATH",
69
+ "LOCATED_IN",
70
+ "WORKPLACE"
71
+ ],
72
+ "PERSON,LOCATION": [
73
+ "PLACE_RESIDES_IN",
74
+ "PLACE_OF_BIRTH",
75
+ "PLACE_OF_DEATH",
76
+ "LOCATED_IN",
77
+ "WORKPLACE"
78
+ ],
79
+ "PERSON,STATE_OR_PROVINCE": [
80
+ "PLACE_RESIDES_IN",
81
+ "PLACE_OF_BIRTH",
82
+ "PLACE_OF_DEATH",
83
+ "LOCATED_IN",
84
+ "WORKPLACE"
85
+ ],
86
+ "PROFESSION,CITY": [
87
+ "PLACE_RESIDES_IN",
88
+ "FOUNDED_BY",
89
+ "PLACE_OF_BIRTH",
90
+ "LOCATED_IN",
91
+ "WORKPLACE",
92
+ "PLACE_OF_DEATH"
93
+ ],
94
+ "PROFESSION,COUNTRY": [
95
+ "PLACE_RESIDES_IN",
96
+ "FOUNDED_BY",
97
+ "PLACE_OF_BIRTH",
98
+ "LOCATED_IN",
99
+ "WORKPLACE",
100
+ "MEMBER_OF",
101
+ "PLACE_OF_DEATH"
102
+ ],
103
+ "PROFESSION,DISTRICT": [
104
+ "PLACE_RESIDES_IN",
105
+ "FOUNDED_BY",
106
+ "PLACE_OF_BIRTH",
107
+ "LOCATED_IN",
108
+ "WORKPLACE",
109
+ "PLACE_OF_DEATH"
110
+ ],
111
+ "PROFESSION,FACILITY": [
112
+ "PLACE_RESIDES_IN",
113
+ "FOUNDED_BY",
114
+ "PLACE_OF_BIRTH",
115
+ "LOCATED_IN",
116
+ "WORKPLACE",
117
+ "PLACE_OF_DEATH"
118
+ ],
119
+ "PROFESSION,LOCATION": [
120
+ "PLACE_RESIDES_IN",
121
+ "FOUNDED_BY",
122
+ "PLACE_OF_BIRTH",
123
+ "LOCATED_IN",
124
+ "WORKPLACE",
125
+ "PLACE_OF_DEATH"
126
+ ],
127
+ "PROFESSION,STATE_OR_PROVINCE": [
128
+ "PLACE_RESIDES_IN",
129
+ "FOUNDED_BY",
130
+ "PLACE_OF_BIRTH",
131
+ "LOCATED_IN",
132
+ "WORKPLACE",
133
+ "PLACE_OF_DEATH"
134
+ ],
135
+ "PERSON,DISEASE": [
136
+ "CAUSE_OF_DEATH",
137
+ "MEDICAL_CONDITION"
138
+ ],
139
+ "PERSON,EVENT": [
140
+ "CAUSE_OF_DEATH",
141
+ "AGENT",
142
+ "INANIMATE_INVOLVED",
143
+ "PARTICIPANT_IN",
144
+ "WORKPLACE",
145
+ "ORGANIZES"
146
+ ],
147
+ "PROFESSION,DISEASE": [
148
+ "CAUSE_OF_DEATH",
149
+ "MEDICAL_CONDITION"
150
+ ],
151
+ "PROFESSION,EVENT": [
152
+ "CAUSE_OF_DEATH",
153
+ "FOUNDED_BY",
154
+ "AGENT",
155
+ "PARTICIPANT_IN",
156
+ "WORKPLACE",
157
+ "ORGANIZES"
158
+ ],
159
+ "NATIONALITY,DISEASE": [
160
+ "CAUSE_OF_DEATH"
161
+ ],
162
+ "NATIONALITY,EVENT": [
163
+ "CAUSE_OF_DEATH",
164
+ "PARTICIPANT_IN",
165
+ "AGENT"
166
+ ],
167
+ "CITY,DATE": [
168
+ "DATE_FOUNDED_IN",
169
+ "DATE_DEFUNCT_IN"
170
+ ],
171
+ "COUNTRY,DATE": [
172
+ "DATE_FOUNDED_IN",
173
+ "DATE_DEFUNCT_IN"
174
+ ],
175
+ "DISTRICT,DATE": [
176
+ "DATE_FOUNDED_IN",
177
+ "DATE_DEFUNCT_IN"
178
+ ],
179
+ "FACILITY,DATE": [
180
+ "DATE_FOUNDED_IN",
181
+ "DATE_DEFUNCT_IN"
182
+ ],
183
+ "EVENT,DATE": [
184
+ "DATE_DEFUNCT_IN",
185
+ "DATE_FOUNDED_IN",
186
+ "START_TIME",
187
+ "END_TIME",
188
+ "POINT_IN_TIME"
189
+ ],
190
+ "ORGANIZATION,DATE": [
191
+ "DATE_FOUNDED_IN",
192
+ "DATE_DEFUNCT_IN"
193
+ ],
194
+ "STATE_OR_PROVINCE,DATE": [
195
+ "DATE_FOUNDED_IN",
196
+ "DATE_DEFUNCT_IN"
197
+ ],
198
+ "WORK_OF_ART,DATE": [
199
+ "DATE_DEFUNCT_IN",
200
+ "DATE_FOUNDED_IN",
201
+ "START_TIME",
202
+ "END_TIME",
203
+ "POINT_IN_TIME",
204
+ "DATE_OF_CREATION"
205
+ ],
206
+ "LOCATION,DATE": [
207
+ "DATE_FOUNDED_IN"
208
+ ],
209
+ "PERSON,DATE": [
210
+ "DATE_OF_BIRTH",
211
+ "DATE_OF_DEATH"
212
+ ],
213
+ "PROFESSION,DATE": [
214
+ "DATE_OF_BIRTH",
215
+ "DATE_OF_DEATH"
216
+ ],
217
+ "LAW,DATE": [
218
+ "DATE_OF_CREATION"
219
+ ],
220
+ "AWARD,DATE": [
221
+ "DATE_OF_CREATION",
222
+ "POINT_IN_TIME"
223
+ ],
224
+ "PRODUCT,DATE": [
225
+ "DATE_OF_CREATION",
226
+ "POINT_IN_TIME"
227
+ ],
228
+ "NATIONALITY,DATE": [
229
+ "DATE_OF_DEATH"
230
+ ],
231
+ "EVENT,TIME": [
232
+ "START_TIME",
233
+ "END_TIME",
234
+ "POINT_IN_TIME"
235
+ ],
236
+ "PENALTY,DATE": [
237
+ "START_TIME",
238
+ "END_TIME",
239
+ "POINT_IN_TIME"
240
+ ],
241
+ "PENALTY,TIME": [
242
+ "START_TIME",
243
+ "END_TIME",
244
+ "POINT_IN_TIME"
245
+ ],
246
+ "CRIME,DATE": [
247
+ "START_TIME",
248
+ "END_TIME",
249
+ "POINT_IN_TIME"
250
+ ],
251
+ "CRIME,TIME": [
252
+ "START_TIME",
253
+ "END_TIME",
254
+ "POINT_IN_TIME"
255
+ ],
256
+ "WORK_OF_ART,TIME": [
257
+ "START_TIME",
258
+ "END_TIME",
259
+ "POINT_IN_TIME"
260
+ ],
261
+ "AWARD,TIME": [
262
+ "POINT_IN_TIME"
263
+ ],
264
+ "PRODUCT,TIME": [
265
+ "POINT_IN_TIME"
266
+ ],
267
+ "NATIONALITY,CITY": [
268
+ "PLACE_RESIDES_IN",
269
+ "PLACE_OF_DEATH"
270
+ ],
271
+ "NATIONALITY,COUNTRY": [
272
+ "PLACE_RESIDES_IN",
273
+ "PLACE_OF_DEATH"
274
+ ],
275
+ "NATIONALITY,DISTRICT": [
276
+ "PLACE_RESIDES_IN",
277
+ "PLACE_OF_DEATH"
278
+ ],
279
+ "NATIONALITY,FACILITY": [
280
+ "PLACE_RESIDES_IN",
281
+ "PLACE_OF_DEATH"
282
+ ],
283
+ "NATIONALITY,LOCATION": [
284
+ "PLACE_RESIDES_IN",
285
+ "PLACE_OF_DEATH"
286
+ ],
287
+ "NATIONALITY,STATE_OR_PROVINCE": [
288
+ "PLACE_RESIDES_IN",
289
+ "PLACE_OF_DEATH"
290
+ ],
291
+ "CITY,CITY": [
292
+ "FOUNDED_BY",
293
+ "LOCATED_IN"
294
+ ],
295
+ "CITY,COUNTRY": [
296
+ "FOUNDED_BY",
297
+ "LOCATED_IN"
298
+ ],
299
+ "CITY,DISTRICT": [
300
+ "FOUNDED_BY",
301
+ "LOCATED_IN"
302
+ ],
303
+ "CITY,FACILITY": [
304
+ "FOUNDED_BY",
305
+ "LOCATED_IN"
306
+ ],
307
+ "CITY,EVENT": [
308
+ "FOUNDED_BY",
309
+ "AGENT",
310
+ "PARTICIPANT_IN",
311
+ "ORGANIZES"
312
+ ],
313
+ "CITY,LOCATION": [
314
+ "FOUNDED_BY",
315
+ "LOCATED_IN"
316
+ ],
317
+ "CITY,ORGANIZATION": [
318
+ "FOUNDED_BY",
319
+ "LOCATED_IN"
320
+ ],
321
+ "CITY,PERSON": [
322
+ "FOUNDED_BY"
323
+ ],
324
+ "CITY,PROFESSION": [
325
+ "FOUNDED_BY"
326
+ ],
327
+ "CITY,STATE_OR_PROVINCE": [
328
+ "FOUNDED_BY",
329
+ "LOCATED_IN"
330
+ ],
331
+ "CITY,FAMILY": [
332
+ "FOUNDED_BY"
333
+ ],
334
+ "COUNTRY,CITY": [
335
+ "FOUNDED_BY",
336
+ "LOCATED_IN"
337
+ ],
338
+ "COUNTRY,COUNTRY": [
339
+ "MEMBER_OF",
340
+ "FOUNDED_BY",
341
+ "LOCATED_IN"
342
+ ],
343
+ "COUNTRY,DISTRICT": [
344
+ "FOUNDED_BY",
345
+ "LOCATED_IN"
346
+ ],
347
+ "COUNTRY,FACILITY": [
348
+ "FOUNDED_BY",
349
+ "LOCATED_IN"
350
+ ],
351
+ "COUNTRY,EVENT": [
352
+ "FOUNDED_BY",
353
+ "AGENT",
354
+ "PARTICIPANT_IN",
355
+ "ORGANIZES"
356
+ ],
357
+ "COUNTRY,LOCATION": [
358
+ "FOUNDED_BY",
359
+ "LOCATED_IN"
360
+ ],
361
+ "COUNTRY,ORGANIZATION": [
362
+ "MEMBER_OF",
363
+ "FOUNDED_BY",
364
+ "LOCATED_IN"
365
+ ],
366
+ "COUNTRY,PERSON": [
367
+ "FOUNDED_BY"
368
+ ],
369
+ "COUNTRY,PROFESSION": [
370
+ "FOUNDED_BY"
371
+ ],
372
+ "COUNTRY,STATE_OR_PROVINCE": [
373
+ "FOUNDED_BY",
374
+ "LOCATED_IN"
375
+ ],
376
+ "COUNTRY,FAMILY": [
377
+ "FOUNDED_BY",
378
+ "MEMBER_OF"
379
+ ],
380
+ "DISTRICT,CITY": [
381
+ "FOUNDED_BY",
382
+ "LOCATED_IN"
383
+ ],
384
+ "DISTRICT,COUNTRY": [
385
+ "FOUNDED_BY",
386
+ "LOCATED_IN"
387
+ ],
388
+ "DISTRICT,DISTRICT": [
389
+ "FOUNDED_BY",
390
+ "LOCATED_IN"
391
+ ],
392
+ "DISTRICT,FACILITY": [
393
+ "FOUNDED_BY",
394
+ "LOCATED_IN"
395
+ ],
396
+ "DISTRICT,EVENT": [
397
+ "FOUNDED_BY",
398
+ "ORGANIZES"
399
+ ],
400
+ "DISTRICT,LOCATION": [
401
+ "FOUNDED_BY",
402
+ "LOCATED_IN"
403
+ ],
404
+ "DISTRICT,ORGANIZATION": [
405
+ "FOUNDED_BY",
406
+ "LOCATED_IN"
407
+ ],
408
+ "DISTRICT,PERSON": [
409
+ "FOUNDED_BY"
410
+ ],
411
+ "DISTRICT,PROFESSION": [
412
+ "FOUNDED_BY"
413
+ ],
414
+ "DISTRICT,STATE_OR_PROVINCE": [
415
+ "FOUNDED_BY",
416
+ "LOCATED_IN"
417
+ ],
418
+ "DISTRICT,FAMILY": [
419
+ "FOUNDED_BY"
420
+ ],
421
+ "FACILITY,CITY": [
422
+ "FOUNDED_BY",
423
+ "LOCATED_IN"
424
+ ],
425
+ "FACILITY,COUNTRY": [
426
+ "FOUNDED_BY",
427
+ "LOCATED_IN"
428
+ ],
429
+ "FACILITY,DISTRICT": [
430
+ "FOUNDED_BY",
431
+ "LOCATED_IN"
432
+ ],
433
+ "FACILITY,FACILITY": [
434
+ "FOUNDED_BY",
435
+ "PART_OF",
436
+ "LOCATED_IN"
437
+ ],
438
+ "FACILITY,EVENT": [
439
+ "INANIMATE_INVOLVED",
440
+ "FOUNDED_BY",
441
+ "PARTICIPANT_IN"
442
+ ],
443
+ "FACILITY,LOCATION": [
444
+ "FOUNDED_BY",
445
+ "LOCATED_IN"
446
+ ],
447
+ "FACILITY,ORGANIZATION": [
448
+ "FOUNDED_BY",
449
+ "PART_OF",
450
+ "LOCATED_IN"
451
+ ],
452
+ "FACILITY,PERSON": [
453
+ "FOUNDED_BY"
454
+ ],
455
+ "FACILITY,PROFESSION": [
456
+ "FOUNDED_BY"
457
+ ],
458
+ "FACILITY,STATE_OR_PROVINCE": [
459
+ "FOUNDED_BY",
460
+ "LOCATED_IN"
461
+ ],
462
+ "FACILITY,FAMILY": [
463
+ "FOUNDED_BY"
464
+ ],
465
+ "EVENT,CITY": [
466
+ "TAKES_PLACE_IN",
467
+ "FOUNDED_BY"
468
+ ],
469
+ "EVENT,COUNTRY": [
470
+ "TAKES_PLACE_IN",
471
+ "FOUNDED_BY"
472
+ ],
473
+ "EVENT,DISTRICT": [
474
+ "TAKES_PLACE_IN",
475
+ "FOUNDED_BY"
476
+ ],
477
+ "EVENT,FACILITY": [
478
+ "TAKES_PLACE_IN",
479
+ "FOUNDED_BY"
480
+ ],
481
+ "EVENT,EVENT": [
482
+ "SUBEVENT_OF",
483
+ "FOUNDED_BY",
484
+ "HAS_CAUSE"
485
+ ],
486
+ "EVENT,LOCATION": [
487
+ "TAKES_PLACE_IN",
488
+ "FOUNDED_BY"
489
+ ],
490
+ "EVENT,ORGANIZATION": [
491
+ "TAKES_PLACE_IN",
492
+ "FOUNDED_BY"
493
+ ],
494
+ "EVENT,PERSON": [
495
+ "FOUNDED_BY"
496
+ ],
497
+ "EVENT,PROFESSION": [
498
+ "FOUNDED_BY"
499
+ ],
500
+ "EVENT,STATE_OR_PROVINCE": [
501
+ "TAKES_PLACE_IN",
502
+ "FOUNDED_BY"
503
+ ],
504
+ "EVENT,FAMILY": [
505
+ "FOUNDED_BY"
506
+ ],
507
+ "LOCATION,CITY": [
508
+ "FOUNDED_BY",
509
+ "LOCATED_IN"
510
+ ],
511
+ "LOCATION,COUNTRY": [
512
+ "FOUNDED_BY",
513
+ "LOCATED_IN"
514
+ ],
515
+ "LOCATION,DISTRICT": [
516
+ "FOUNDED_BY",
517
+ "LOCATED_IN"
518
+ ],
519
+ "LOCATION,FACILITY": [
520
+ "FOUNDED_BY",
521
+ "LOCATED_IN"
522
+ ],
523
+ "LOCATION,EVENT": [
524
+ "FOUNDED_BY"
525
+ ],
526
+ "LOCATION,LOCATION": [
527
+ "FOUNDED_BY",
528
+ "LOCATED_IN"
529
+ ],
530
+ "LOCATION,ORGANIZATION": [
531
+ "FOUNDED_BY",
532
+ "LOCATED_IN"
533
+ ],
534
+ "LOCATION,PERSON": [
535
+ "FOUNDED_BY"
536
+ ],
537
+ "LOCATION,PROFESSION": [
538
+ "FOUNDED_BY"
539
+ ],
540
+ "LOCATION,STATE_OR_PROVINCE": [
541
+ "FOUNDED_BY",
542
+ "LOCATED_IN"
543
+ ],
544
+ "LOCATION,FAMILY": [
545
+ "FOUNDED_BY"
546
+ ],
547
+ "ORGANIZATION,CITY": [
548
+ "FOUNDED_BY",
549
+ "HEADQUARTERED_IN",
550
+ "LOCATED_IN"
551
+ ],
552
+ "ORGANIZATION,COUNTRY": [
553
+ "MEMBER_OF",
554
+ "FOUNDED_BY",
555
+ "HEADQUARTERED_IN",
556
+ "LOCATED_IN"
557
+ ],
558
+ "ORGANIZATION,DISTRICT": [
559
+ "FOUNDED_BY",
560
+ "HEADQUARTERED_IN",
561
+ "LOCATED_IN"
562
+ ],
563
+ "ORGANIZATION,FACILITY": [
564
+ "FOUNDED_BY",
565
+ "PART_OF",
566
+ "HEADQUARTERED_IN",
567
+ "LOCATED_IN"
568
+ ],
569
+ "ORGANIZATION,EVENT": [
570
+ "FOUNDED_BY",
571
+ "AGENT",
572
+ "PARTICIPANT_IN",
573
+ "ORGANIZES"
574
+ ],
575
+ "ORGANIZATION,LOCATION": [
576
+ "FOUNDED_BY",
577
+ "HEADQUARTERED_IN",
578
+ "LOCATED_IN"
579
+ ],
580
+ "ORGANIZATION,ORGANIZATION": [
581
+ "MEMBER_OF",
582
+ "FOUNDED_BY",
583
+ "PART_OF",
584
+ "LOCATED_IN"
585
+ ],
586
+ "ORGANIZATION,PERSON": [
587
+ "FOUNDED_BY"
588
+ ],
589
+ "ORGANIZATION,PROFESSION": [
590
+ "FOUNDED_BY"
591
+ ],
592
+ "ORGANIZATION,STATE_OR_PROVINCE": [
593
+ "FOUNDED_BY",
594
+ "HEADQUARTERED_IN",
595
+ "LOCATED_IN"
596
+ ],
597
+ "ORGANIZATION,FAMILY": [
598
+ "FOUNDED_BY",
599
+ "MEMBER_OF"
600
+ ],
601
+ "STATE_OR_PROVINCE,CITY": [
602
+ "FOUNDED_BY",
603
+ "LOCATED_IN"
604
+ ],
605
+ "STATE_OR_PROVINCE,COUNTRY": [
606
+ "FOUNDED_BY",
607
+ "LOCATED_IN"
608
+ ],
609
+ "STATE_OR_PROVINCE,DISTRICT": [
610
+ "FOUNDED_BY",
611
+ "LOCATED_IN"
612
+ ],
613
+ "STATE_OR_PROVINCE,FACILITY": [
614
+ "FOUNDED_BY",
615
+ "LOCATED_IN"
616
+ ],
617
+ "STATE_OR_PROVINCE,EVENT": [
618
+ "FOUNDED_BY",
619
+ "AGENT",
620
+ "PARTICIPANT_IN",
621
+ "ORGANIZES"
622
+ ],
623
+ "STATE_OR_PROVINCE,LOCATION": [
624
+ "FOUNDED_BY",
625
+ "LOCATED_IN"
626
+ ],
627
+ "STATE_OR_PROVINCE,ORGANIZATION": [
628
+ "FOUNDED_BY",
629
+ "LOCATED_IN"
630
+ ],
631
+ "STATE_OR_PROVINCE,PERSON": [
632
+ "FOUNDED_BY"
633
+ ],
634
+ "STATE_OR_PROVINCE,PROFESSION": [
635
+ "FOUNDED_BY"
636
+ ],
637
+ "STATE_OR_PROVINCE,STATE_OR_PROVINCE": [
638
+ "FOUNDED_BY",
639
+ "LOCATED_IN"
640
+ ],
641
+ "STATE_OR_PROVINCE,FAMILY": [
642
+ "FOUNDED_BY"
643
+ ],
644
+ "PROFESSION,ORGANIZATION": [
645
+ "FOUNDED_BY",
646
+ "SCHOOLS_ATTENDED",
647
+ "LOCATED_IN",
648
+ "MEMBER_OF",
649
+ "WORKPLACE"
650
+ ],
651
+ "PROFESSION,PERSON": [
652
+ "SPOUSE",
653
+ "PARENT_OF",
654
+ "FOUNDED_BY",
655
+ "SUBORDINATE_OF",
656
+ "SIBLING"
657
+ ],
658
+ "PROFESSION,PROFESSION": [
659
+ "SPOUSE",
660
+ "PARENT_OF",
661
+ "FOUNDED_BY",
662
+ "SUBORDINATE_OF",
663
+ "SIBLING"
664
+ ],
665
+ "PROFESSION,FAMILY": [
666
+ "FOUNDED_BY",
667
+ "MEMBER_OF"
668
+ ],
669
+ "PERSON,IDEOLOGY": [
670
+ "IDEOLOGY_OF",
671
+ "MEMBER_OF",
672
+ "WORKPLACE"
673
+ ],
674
+ "ORGANIZATION,IDEOLOGY": [
675
+ "IDEOLOGY_OF",
676
+ "MEMBER_OF"
677
+ ],
678
+ "PROFESSION,IDEOLOGY": [
679
+ "IDEOLOGY_OF",
680
+ "MEMBER_OF",
681
+ "WORKPLACE"
682
+ ],
683
+ "COUNTRY,IDEOLOGY": [
684
+ "IDEOLOGY_OF",
685
+ "MEMBER_OF"
686
+ ],
687
+ "FACILITY,IDEOLOGY": [
688
+ "IDEOLOGY_OF"
689
+ ],
690
+ "NATIONALITY,IDEOLOGY": [
691
+ "IDEOLOGY_OF"
692
+ ],
693
+ "EVENT,IDEOLOGY": [
694
+ "IDEOLOGY_OF"
695
+ ],
696
+ "PERSON,ORGANIZATION": [
697
+ "MEMBER_OF",
698
+ "SCHOOLS_ATTENDED",
699
+ "WORKPLACE",
700
+ "LOCATED_IN"
701
+ ],
702
+ "PRODUCT,CITY": [
703
+ "LOCATED_IN"
704
+ ],
705
+ "PRODUCT,COUNTRY": [
706
+ "LOCATED_IN"
707
+ ],
708
+ "PRODUCT,DISTRICT": [
709
+ "LOCATED_IN"
710
+ ],
711
+ "PRODUCT,FACILITY": [
712
+ "PART_OF",
713
+ "LOCATED_IN"
714
+ ],
715
+ "PRODUCT,LOCATION": [
716
+ "LOCATED_IN"
717
+ ],
718
+ "PRODUCT,ORGANIZATION": [
719
+ "PART_OF",
720
+ "LOCATED_IN"
721
+ ],
722
+ "PRODUCT,STATE_OR_PROVINCE": [
723
+ "LOCATED_IN"
724
+ ],
725
+ "WORK_OF_ART,CITY": [
726
+ "LOCATED_IN"
727
+ ],
728
+ "WORK_OF_ART,COUNTRY": [
729
+ "LOCATED_IN"
730
+ ],
731
+ "WORK_OF_ART,DISTRICT": [
732
+ "LOCATED_IN"
733
+ ],
734
+ "WORK_OF_ART,FACILITY": [
735
+ "PART_OF",
736
+ "LOCATED_IN"
737
+ ],
738
+ "WORK_OF_ART,LOCATION": [
739
+ "LOCATED_IN"
740
+ ],
741
+ "WORK_OF_ART,ORGANIZATION": [
742
+ "PART_OF",
743
+ "LOCATED_IN"
744
+ ],
745
+ "WORK_OF_ART,STATE_OR_PROVINCE": [
746
+ "LOCATED_IN"
747
+ ],
748
+ "PERSON,PERSON": [
749
+ "PARENT_OF",
750
+ "SPOUSE",
751
+ "SIBLING",
752
+ "SUBORDINATE_OF"
753
+ ],
754
+ "PERSON,PROFESSION": [
755
+ "SPOUSE",
756
+ "PARENT_OF",
757
+ "SUBORDINATE_OF",
758
+ "WORKS_AS",
759
+ "SIBLING"
760
+ ],
761
+ "PERSON,FAMILY": [
762
+ "MEMBER_OF"
763
+ ],
764
+ "PERSON,NATIONALITY": [
765
+ "PARENT_OF"
766
+ ],
767
+ "PROFESSION,NATIONALITY": [
768
+ "PARENT_OF"
769
+ ],
770
+ "NATIONALITY,PERSON": [
771
+ "PARENT_OF"
772
+ ],
773
+ "NATIONALITY,PROFESSION": [
774
+ "PARENT_OF"
775
+ ],
776
+ "NATIONALITY,NATIONALITY": [
777
+ "PARENT_OF"
778
+ ],
779
+ "FAMILY,CITY": [
780
+ "PLACE_RESIDES_IN"
781
+ ],
782
+ "FAMILY,COUNTRY": [
783
+ "PLACE_RESIDES_IN"
784
+ ],
785
+ "FAMILY,DISTRICT": [
786
+ "PLACE_RESIDES_IN"
787
+ ],
788
+ "FAMILY,FACILITY": [
789
+ "PLACE_RESIDES_IN"
790
+ ],
791
+ "FAMILY,LOCATION": [
792
+ "PLACE_RESIDES_IN"
793
+ ],
794
+ "FAMILY,STATE_OR_PROVINCE": [
795
+ "PLACE_RESIDES_IN"
796
+ ],
797
+ "<ENTITY>,MONEY": [
798
+ "PRICE_OF"
799
+ ],
800
+ "CITY,<ENTITY>": [
801
+ "PRODUCES"
802
+ ],
803
+ "COUNTRY,<ENTITY>": [
804
+ "PRODUCES"
805
+ ],
806
+ "DISTRICT,<ENTITY>": [
807
+ "PRODUCES"
808
+ ],
809
+ "ORGANIZATION,<ENTITY>": [
810
+ "PRODUCES"
811
+ ],
812
+ "STATE_OR_PROVINCE,<ENTITY>": [
813
+ "PRODUCES"
814
+ ],
815
+ "PERSON,RELIGION": [
816
+ "RELIGION_OF"
817
+ ],
818
+ "ORGANIZATION,RELIGION": [
819
+ "RELIGION_OF"
820
+ ],
821
+ "PROFESSION,RELIGION": [
822
+ "RELIGION_OF"
823
+ ],
824
+ "COUNTRY,RELIGION": [
825
+ "RELIGION_OF"
826
+ ],
827
+ "FACILITY,RELIGION": [
828
+ "RELIGION_OF"
829
+ ],
830
+ "NATIONALITY,RELIGION": [
831
+ "RELIGION_OF"
832
+ ],
833
+ "EVENT,RELIGION": [
834
+ "RELIGION_OF"
835
+ ],
836
+ "NATIONALITY,ORGANIZATION": [
837
+ "SCHOOLS_ATTENDED"
838
+ ],
839
+ "CRIME,CITY": [
840
+ "TAKES_PLACE_IN"
841
+ ],
842
+ "CRIME,COUNTRY": [
843
+ "TAKES_PLACE_IN"
844
+ ],
845
+ "CRIME,DISTRICT": [
846
+ "TAKES_PLACE_IN"
847
+ ],
848
+ "CRIME,ORGANIZATION": [
849
+ "TAKES_PLACE_IN"
850
+ ],
851
+ "CRIME,STATE_OR_PROVINCE": [
852
+ "TAKES_PLACE_IN"
853
+ ],
854
+ "CRIME,FACILITY": [
855
+ "TAKES_PLACE_IN"
856
+ ],
857
+ "CRIME,LOCATION": [
858
+ "TAKES_PLACE_IN"
859
+ ],
860
+ "PENALTY,CITY": [
861
+ "TAKES_PLACE_IN"
862
+ ],
863
+ "PENALTY,COUNTRY": [
864
+ "TAKES_PLACE_IN"
865
+ ],
866
+ "PENALTY,DISTRICT": [
867
+ "TAKES_PLACE_IN"
868
+ ],
869
+ "PENALTY,ORGANIZATION": [
870
+ "TAKES_PLACE_IN"
871
+ ],
872
+ "PENALTY,STATE_OR_PROVINCE": [
873
+ "TAKES_PLACE_IN"
874
+ ],
875
+ "PENALTY,FACILITY": [
876
+ "TAKES_PLACE_IN"
877
+ ],
878
+ "PENALTY,LOCATION": [
879
+ "TAKES_PLACE_IN"
880
+ ],
881
+ "PERSON,CRIME": [
882
+ "PARTICIPANT_IN",
883
+ "INANIMATE_INVOLVED",
884
+ "CONVICTED_OF"
885
+ ],
886
+ "PROFESSION,CRIME": [
887
+ "PARTICIPANT_IN",
888
+ "CONVICTED_OF"
889
+ ],
890
+ "ORGANIZATION,CRIME": [
891
+ "PARTICIPANT_IN",
892
+ "CONVICTED_OF"
893
+ ],
894
+ "FAMILY,CRIME": [
895
+ "PARTICIPANT_IN",
896
+ "CONVICTED_OF"
897
+ ],
898
+ "NATIONALITY,CRIME": [
899
+ "PARTICIPANT_IN",
900
+ "CONVICTED_OF"
901
+ ],
902
+ "COUNTRY,CRIME": [
903
+ "PARTICIPANT_IN",
904
+ "CONVICTED_OF"
905
+ ],
906
+ "PERSON,PENALTY": [
907
+ "PENALIZED_AS",
908
+ "PARTICIPANT_IN",
909
+ "INANIMATE_INVOLVED"
910
+ ],
911
+ "PROFESSION,PENALTY": [
912
+ "PENALIZED_AS",
913
+ "PARTICIPANT_IN"
914
+ ],
915
+ "ORGANIZATION,PENALTY": [
916
+ "PENALIZED_AS",
917
+ "PARTICIPANT_IN"
918
+ ],
919
+ "FAMILY,PENALTY": [
920
+ "PENALIZED_AS",
921
+ "PARTICIPANT_IN"
922
+ ],
923
+ "NATIONALITY,PENALTY": [
924
+ "PENALIZED_AS",
925
+ "PARTICIPANT_IN"
926
+ ],
927
+ "COUNTRY,PENALTY": [
928
+ "PENALIZED_AS",
929
+ "PARTICIPANT_IN"
930
+ ],
931
+ "ORGANIZATION,WORK_OF_ART": [
932
+ "PARTICIPANT_IN",
933
+ "PART_OF"
934
+ ],
935
+ "ORGANIZATION,LAW": [
936
+ "PART_OF"
937
+ ],
938
+ "ORGANIZATION,PRODUCT": [
939
+ "PART_OF"
940
+ ],
941
+ "WORK_OF_ART,WORK_OF_ART": [
942
+ "PARTICIPANT_IN",
943
+ "PART_OF",
944
+ "INANIMATE_INVOLVED"
945
+ ],
946
+ "WORK_OF_ART,LAW": [
947
+ "PART_OF"
948
+ ],
949
+ "WORK_OF_ART,PRODUCT": [
950
+ "PART_OF"
951
+ ],
952
+ "LAW,ORGANIZATION": [
953
+ "PART_OF"
954
+ ],
955
+ "LAW,WORK_OF_ART": [
956
+ "PART_OF",
957
+ "INANIMATE_INVOLVED"
958
+ ],
959
+ "LAW,LAW": [
960
+ "PART_OF"
961
+ ],
962
+ "LAW,FACILITY": [
963
+ "PART_OF"
964
+ ],
965
+ "LAW,PRODUCT": [
966
+ "PART_OF"
967
+ ],
968
+ "LAW,AWARD": [
969
+ "PART_OF"
970
+ ],
971
+ "FACILITY,WORK_OF_ART": [
972
+ "PARTICIPANT_IN",
973
+ "PART_OF",
974
+ "INANIMATE_INVOLVED"
975
+ ],
976
+ "FACILITY,LAW": [
977
+ "PART_OF"
978
+ ],
979
+ "FACILITY,PRODUCT": [
980
+ "PART_OF"
981
+ ],
982
+ "FACILITY,AWARD": [
983
+ "PART_OF"
984
+ ],
985
+ "PRODUCT,WORK_OF_ART": [
986
+ "PART_OF",
987
+ "INANIMATE_INVOLVED"
988
+ ],
989
+ "PRODUCT,LAW": [
990
+ "PART_OF"
991
+ ],
992
+ "PRODUCT,PRODUCT": [
993
+ "PART_OF"
994
+ ],
995
+ "PRODUCT,AWARD": [
996
+ "PART_OF"
997
+ ],
998
+ "AWARD,ORGANIZATION": [
999
+ "PART_OF"
1000
+ ],
1001
+ "AWARD,WORK_OF_ART": [
1002
+ "PARTICIPANT_IN",
1003
+ "PART_OF",
1004
+ "INANIMATE_INVOLVED"
1005
+ ],
1006
+ "AWARD,LAW": [
1007
+ "PART_OF",
1008
+ "HAS_CAUSE"
1009
+ ],
1010
+ "AWARD,FACILITY": [
1011
+ "PART_OF"
1012
+ ],
1013
+ "AWARD,PRODUCT": [
1014
+ "PART_OF"
1015
+ ],
1016
+ "AWARD,AWARD": [
1017
+ "PART_OF"
1018
+ ],
1019
+ "EVENT,CRIME": [
1020
+ "HAS_CAUSE"
1021
+ ],
1022
+ "EVENT,PENALTY": [
1023
+ "HAS_CAUSE"
1024
+ ],
1025
+ "EVENT,LAW": [
1026
+ "HAS_CAUSE"
1027
+ ],
1028
+ "EVENT,DISEASE": [
1029
+ "HAS_CAUSE"
1030
+ ],
1031
+ "CRIME,EVENT": [
1032
+ "HAS_CAUSE"
1033
+ ],
1034
+ "CRIME,CRIME": [
1035
+ "HAS_CAUSE"
1036
+ ],
1037
+ "CRIME,PENALTY": [
1038
+ "HAS_CAUSE"
1039
+ ],
1040
+ "CRIME,LAW": [
1041
+ "HAS_CAUSE"
1042
+ ],
1043
+ "CRIME,DISEASE": [
1044
+ "HAS_CAUSE"
1045
+ ],
1046
+ "PENALTY,EVENT": [
1047
+ "HAS_CAUSE"
1048
+ ],
1049
+ "PENALTY,CRIME": [
1050
+ "HAS_CAUSE"
1051
+ ],
1052
+ "PENALTY,PENALTY": [
1053
+ "HAS_CAUSE"
1054
+ ],
1055
+ "PENALTY,LAW": [
1056
+ "HAS_CAUSE"
1057
+ ],
1058
+ "PENALTY,DISEASE": [
1059
+ "HAS_CAUSE"
1060
+ ],
1061
+ "AWARD,EVENT": [
1062
+ "PARTICIPANT_IN",
1063
+ "INANIMATE_INVOLVED",
1064
+ "HAS_CAUSE"
1065
+ ],
1066
+ "AWARD,CRIME": [
1067
+ "PARTICIPANT_IN",
1068
+ "INANIMATE_INVOLVED",
1069
+ "HAS_CAUSE"
1070
+ ],
1071
+ "AWARD,PENALTY": [
1072
+ "PARTICIPANT_IN",
1073
+ "INANIMATE_INVOLVED",
1074
+ "HAS_CAUSE"
1075
+ ],
1076
+ "AWARD,DISEASE": [
1077
+ "HAS_CAUSE"
1078
+ ],
1079
+ "DISEASE,EVENT": [
1080
+ "HAS_CAUSE"
1081
+ ],
1082
+ "DISEASE,CRIME": [
1083
+ "HAS_CAUSE"
1084
+ ],
1085
+ "DISEASE,PENALTY": [
1086
+ "HAS_CAUSE"
1087
+ ],
1088
+ "DISEASE,LAW": [
1089
+ "HAS_CAUSE"
1090
+ ],
1091
+ "DISEASE,DISEASE": [
1092
+ "HAS_CAUSE"
1093
+ ],
1094
+ "FAMILY,EVENT": [
1095
+ "PARTICIPANT_IN",
1096
+ "AGENT"
1097
+ ],
1098
+ "IDEOLOGY,EVENT": [
1099
+ "PARTICIPANT_IN",
1100
+ "AGENT"
1101
+ ],
1102
+ "RELIGION,EVENT": [
1103
+ "PARTICIPANT_IN",
1104
+ "AGENT"
1105
+ ],
1106
+ "PERSON,WORK_OF_ART": [
1107
+ "PARTICIPANT_IN",
1108
+ "INANIMATE_INVOLVED"
1109
+ ],
1110
+ "PROFESSION,WORK_OF_ART": [
1111
+ "PARTICIPANT_IN"
1112
+ ],
1113
+ "CITY,WORK_OF_ART": [
1114
+ "PARTICIPANT_IN"
1115
+ ],
1116
+ "CITY,CRIME": [
1117
+ "PARTICIPANT_IN"
1118
+ ],
1119
+ "CITY,PENALTY": [
1120
+ "PARTICIPANT_IN"
1121
+ ],
1122
+ "COUNTRY,WORK_OF_ART": [
1123
+ "PARTICIPANT_IN"
1124
+ ],
1125
+ "STATE_OR_PROVINCE,WORK_OF_ART": [
1126
+ "PARTICIPANT_IN"
1127
+ ],
1128
+ "STATE_OR_PROVINCE,CRIME": [
1129
+ "PARTICIPANT_IN"
1130
+ ],
1131
+ "STATE_OR_PROVINCE,PENALTY": [
1132
+ "PARTICIPANT_IN"
1133
+ ],
1134
+ "FACILITY,CRIME": [
1135
+ "PARTICIPANT_IN",
1136
+ "INANIMATE_INVOLVED"
1137
+ ],
1138
+ "FACILITY,PENALTY": [
1139
+ "PARTICIPANT_IN",
1140
+ "INANIMATE_INVOLVED"
1141
+ ],
1142
+ "WORK_OF_ART,EVENT": [
1143
+ "PARTICIPANT_IN",
1144
+ "INANIMATE_INVOLVED"
1145
+ ],
1146
+ "WORK_OF_ART,CRIME": [
1147
+ "PARTICIPANT_IN",
1148
+ "INANIMATE_INVOLVED"
1149
+ ],
1150
+ "WORK_OF_ART,PENALTY": [
1151
+ "PARTICIPANT_IN",
1152
+ "INANIMATE_INVOLVED"
1153
+ ],
1154
+ "FAMILY,WORK_OF_ART": [
1155
+ "PARTICIPANT_IN"
1156
+ ],
1157
+ "NATIONALITY,WORK_OF_ART": [
1158
+ "PARTICIPANT_IN"
1159
+ ],
1160
+ "IDEOLOGY,WORK_OF_ART": [
1161
+ "PARTICIPANT_IN"
1162
+ ],
1163
+ "IDEOLOGY,CRIME": [
1164
+ "PARTICIPANT_IN"
1165
+ ],
1166
+ "IDEOLOGY,PENALTY": [
1167
+ "PARTICIPANT_IN"
1168
+ ],
1169
+ "RELIGION,WORK_OF_ART": [
1170
+ "PARTICIPANT_IN"
1171
+ ],
1172
+ "RELIGION,CRIME": [
1173
+ "PARTICIPANT_IN"
1174
+ ],
1175
+ "RELIGION,PENALTY": [
1176
+ "PARTICIPANT_IN"
1177
+ ],
1178
+ "PRODUCT,EVENT": [
1179
+ "INANIMATE_INVOLVED"
1180
+ ],
1181
+ "PRODUCT,CRIME": [
1182
+ "INANIMATE_INVOLVED"
1183
+ ],
1184
+ "PRODUCT,PENALTY": [
1185
+ "INANIMATE_INVOLVED"
1186
+ ],
1187
+ "LAW,EVENT": [
1188
+ "INANIMATE_INVOLVED"
1189
+ ],
1190
+ "LAW,CRIME": [
1191
+ "INANIMATE_INVOLVED"
1192
+ ],
1193
+ "LAW,PENALTY": [
1194
+ "INANIMATE_INVOLVED"
1195
+ ],
1196
+ "MONEY,EVENT": [
1197
+ "INANIMATE_INVOLVED"
1198
+ ],
1199
+ "MONEY,WORK_OF_ART": [
1200
+ "INANIMATE_INVOLVED"
1201
+ ],
1202
+ "MONEY,CRIME": [
1203
+ "INANIMATE_INVOLVED"
1204
+ ],
1205
+ "MONEY,PENALTY": [
1206
+ "INANIMATE_INVOLVED"
1207
+ ],
1208
+ "PERSON,MONEY": [
1209
+ "INCOME",
1210
+ "EXPENDITURE"
1211
+ ],
1212
+ "PROFESSION,MONEY": [
1213
+ "INCOME",
1214
+ "EXPENDITURE"
1215
+ ],
1216
+ "CITY,MONEY": [
1217
+ "INCOME",
1218
+ "EXPENDITURE"
1219
+ ],
1220
+ "COUNTRY,MONEY": [
1221
+ "INCOME",
1222
+ "EXPENDITURE"
1223
+ ],
1224
+ "DISTRICT,MONEY": [
1225
+ "INCOME",
1226
+ "EXPENDITURE"
1227
+ ],
1228
+ "ORGANIZATION,MONEY": [
1229
+ "INCOME",
1230
+ "EXPENDITURE"
1231
+ ],
1232
+ "FAMILY,MONEY": [
1233
+ "INCOME",
1234
+ "EXPENDITURE"
1235
+ ],
1236
+ "STATE_OR_PROVINCE,MONEY": [
1237
+ "INCOME",
1238
+ "EXPENDITURE"
1239
+ ],
1240
+ "NATIONALITY,MONEY": [
1241
+ "INCOME",
1242
+ "EXPENDITURE"
1243
+ ]
1244
+ }
special_tokens_map.json ADDED
@@ -0,0 +1,67 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ {
4
+ "content": "<e1>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false
9
+ },
10
+ {
11
+ "content": "</e1>",
12
+ "lstrip": false,
13
+ "normalized": false,
14
+ "rstrip": false,
15
+ "single_word": false
16
+ },
17
+ {
18
+ "content": "<e2>",
19
+ "lstrip": false,
20
+ "normalized": false,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ },
24
+ {
25
+ "content": "</e2>",
26
+ "lstrip": false,
27
+ "normalized": false,
28
+ "rstrip": false,
29
+ "single_word": false
30
+ }
31
+ ],
32
+ "cls_token": {
33
+ "content": "[CLS]",
34
+ "lstrip": false,
35
+ "normalized": false,
36
+ "rstrip": false,
37
+ "single_word": false
38
+ },
39
+ "mask_token": {
40
+ "content": "[MASK]",
41
+ "lstrip": false,
42
+ "normalized": false,
43
+ "rstrip": false,
44
+ "single_word": false
45
+ },
46
+ "pad_token": {
47
+ "content": "[PAD]",
48
+ "lstrip": false,
49
+ "normalized": false,
50
+ "rstrip": false,
51
+ "single_word": false
52
+ },
53
+ "sep_token": {
54
+ "content": "[SEP]",
55
+ "lstrip": false,
56
+ "normalized": false,
57
+ "rstrip": false,
58
+ "single_word": false
59
+ },
60
+ "unk_token": {
61
+ "content": "[UNK]",
62
+ "lstrip": false,
63
+ "normalized": false,
64
+ "rstrip": false,
65
+ "single_word": false
66
+ }
67
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,96 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "[PAD]",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "100": {
12
+ "content": "[UNK]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "101": {
20
+ "content": "[CLS]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "102": {
28
+ "content": "[SEP]",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "103": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ },
43
+ "119547": {
44
+ "content": "<e1>",
45
+ "lstrip": false,
46
+ "normalized": false,
47
+ "rstrip": false,
48
+ "single_word": false,
49
+ "special": true
50
+ },
51
+ "119548": {
52
+ "content": "</e1>",
53
+ "lstrip": false,
54
+ "normalized": false,
55
+ "rstrip": false,
56
+ "single_word": false,
57
+ "special": true
58
+ },
59
+ "119549": {
60
+ "content": "<e2>",
61
+ "lstrip": false,
62
+ "normalized": false,
63
+ "rstrip": false,
64
+ "single_word": false,
65
+ "special": true
66
+ },
67
+ "119550": {
68
+ "content": "</e2>",
69
+ "lstrip": false,
70
+ "normalized": false,
71
+ "rstrip": false,
72
+ "single_word": false,
73
+ "special": true
74
+ }
75
+ },
76
+ "additional_special_tokens": [
77
+ "<e1>",
78
+ "</e1>",
79
+ "<e2>",
80
+ "</e2>"
81
+ ],
82
+ "clean_up_tokenization_spaces": true,
83
+ "cls_token": "[CLS]",
84
+ "do_basic_tokenize": true,
85
+ "do_lower_case": false,
86
+ "extra_special_tokens": {},
87
+ "mask_token": "[MASK]",
88
+ "model_max_length": 1000000000000000019884624838656,
89
+ "never_split": null,
90
+ "pad_token": "[PAD]",
91
+ "sep_token": "[SEP]",
92
+ "strip_accents": null,
93
+ "tokenize_chinese_chars": true,
94
+ "tokenizer_class": "BertTokenizer",
95
+ "unk_token": "[UNK]"
96
+ }
vocab.txt ADDED
The diff for this file is too large to render. See raw diff