Update ner_tool.py
Browse files- ner_tool.py +241 -45
ner_tool.py
CHANGED
@@ -1,57 +1,253 @@
|
|
1 |
-
|
2 |
-
|
3 |
-
from
|
4 |
-
from transformers import Tool
|
5 |
|
6 |
class NamedEntityRecognitionTool(Tool):
|
7 |
name = "ner_tool"
|
8 |
-
description = "
|
9 |
-
|
10 |
-
|
11 |
-
|
12 |
-
|
13 |
-
|
14 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
15 |
|
16 |
-
|
17 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
18 |
|
19 |
-
|
20 |
-
|
21 |
-
|
22 |
-
|
23 |
-
|
24 |
-
current_label = None
|
25 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
26 |
for entity in entities:
|
|
|
27 |
label = entity.get("entity", "UNKNOWN")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
28 |
word = entity.get("word", "")
|
29 |
-
|
30 |
-
|
31 |
-
|
32 |
-
|
33 |
-
|
34 |
-
|
35 |
-
|
36 |
-
|
37 |
-
|
38 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
39 |
current_label = label
|
|
|
40 |
else:
|
41 |
-
|
42 |
-
|
43 |
-
|
44 |
-
|
45 |
-
|
46 |
-
|
47 |
-
|
48 |
-
|
49 |
-
|
50 |
-
|
51 |
-
|
52 |
-
|
53 |
-
|
54 |
-
|
55 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
56 |
|
57 |
-
|
|
|
|
|
|
|
|
1 |
+
import os
|
2 |
+
from typing import Dict, List, Any, Optional, Union
|
3 |
+
from smolagents import Tool
|
|
|
4 |
|
5 |
class NamedEntityRecognitionTool(Tool):
|
6 |
name = "ner_tool"
|
7 |
+
description = """
|
8 |
+
Identifies and labels named entities in text using customizable NER models.
|
9 |
+
Can recognize entities such as persons, organizations, locations, dates, etc.
|
10 |
+
Returns a structured analysis of all entities found in the input text.
|
11 |
+
"""
|
12 |
+
inputs = {
|
13 |
+
"text": {
|
14 |
+
"type": "string",
|
15 |
+
"description": "The text to analyze for named entities",
|
16 |
+
},
|
17 |
+
"model": {
|
18 |
+
"type": "string",
|
19 |
+
"description": "The NER model to use (default: 'dslim/bert-base-NER')",
|
20 |
+
"nullable": True
|
21 |
+
},
|
22 |
+
"aggregation": {
|
23 |
+
"type": "string",
|
24 |
+
"description": "How to aggregate entities: 'simple' (just list), 'grouped' (by label), or 'detailed' (with confidence scores)",
|
25 |
+
"nullable": True
|
26 |
+
},
|
27 |
+
"min_score": {
|
28 |
+
"type": "number",
|
29 |
+
"description": "Minimum confidence score threshold (0.0-1.0) for including entities",
|
30 |
+
"nullable": True
|
31 |
+
}
|
32 |
+
}
|
33 |
+
output_type = "string"
|
34 |
+
|
35 |
+
def __init__(self):
|
36 |
+
"""Initialize the NER Tool with default settings."""
|
37 |
+
super().__init__()
|
38 |
+
self.default_model = "dslim/bert-base-NER"
|
39 |
+
self.available_models = {
|
40 |
+
"dslim/bert-base-NER": "Standard NER (English)",
|
41 |
+
"jean-baptiste/camembert-ner": "French NER",
|
42 |
+
"Davlan/bert-base-multilingual-cased-ner-hrl": "Multilingual NER",
|
43 |
+
"Babelscape/wikineural-multilingual-ner": "WikiNeural Multilingual NER",
|
44 |
+
"flair/ner-english-ontonotes-large": "OntoNotes English (fine-grained)",
|
45 |
+
"elastic/distilbert-base-cased-finetuned-conll03-english": "CoNLL (fast)"
|
46 |
+
}
|
47 |
+
self.entity_colors = {
|
48 |
+
"PER": "🟥 Person",
|
49 |
+
"PERSON": "🟥 Person",
|
50 |
+
"LOC": "🟨 Location",
|
51 |
+
"LOCATION": "🟨 Location",
|
52 |
+
"GPE": "🟨 Location",
|
53 |
+
"ORG": "🟦 Organization",
|
54 |
+
"ORGANIZATION": "🟦 Organization",
|
55 |
+
"MISC": "🟩 Miscellaneous",
|
56 |
+
"DATE": "🟪 Date",
|
57 |
+
"TIME": "🟪 Time",
|
58 |
+
"MONEY": "💰 Money",
|
59 |
+
"PERCENT": "📊 Percentage",
|
60 |
+
"PRODUCT": "🛒 Product",
|
61 |
+
"EVENT": "🎫 Event",
|
62 |
+
"WORK_OF_ART": "🎨 Work of Art",
|
63 |
+
"LAW": "⚖️ Law",
|
64 |
+
"LANGUAGE": "🗣️ Language",
|
65 |
+
"FAC": "🏢 Facility"
|
66 |
+
}
|
67 |
+
# Pipeline will be lazily loaded
|
68 |
+
self._pipeline = None
|
69 |
|
70 |
+
def _load_pipeline(self, model_name: str):
|
71 |
+
"""Load the NER pipeline with the specified model."""
|
72 |
+
try:
|
73 |
+
from transformers import pipeline
|
74 |
+
self._pipeline = pipeline("ner", model=model_name, aggregation_strategy="simple")
|
75 |
+
return True
|
76 |
+
except Exception as e:
|
77 |
+
print(f"Error loading model {model_name}: {str(e)}")
|
78 |
+
try:
|
79 |
+
# Fall back to default model
|
80 |
+
from transformers import pipeline
|
81 |
+
self._pipeline = pipeline("ner", model=self.default_model, aggregation_strategy="simple")
|
82 |
+
return True
|
83 |
+
except Exception as fallback_error:
|
84 |
+
print(f"Error loading fallback model: {str(fallback_error)}")
|
85 |
+
return False
|
86 |
|
87 |
+
def _get_friendly_label(self, label: str) -> str:
|
88 |
+
"""Convert technical entity labels to friendly descriptions with color indicators."""
|
89 |
+
# Strip B- or I- prefixes that indicate beginning or inside of entity
|
90 |
+
clean_label = label.replace("B-", "").replace("I-", "")
|
91 |
+
return self.entity_colors.get(clean_label, f"🔷 {clean_label}")
|
|
|
92 |
|
93 |
+
def forward(self, text: str, model: str = None, aggregation: str = None, min_score: float = None) -> str:
|
94 |
+
"""
|
95 |
+
Perform Named Entity Recognition on the input text.
|
96 |
+
|
97 |
+
Args:
|
98 |
+
text: The text to analyze
|
99 |
+
model: NER model to use (default: dslim/bert-base-NER)
|
100 |
+
aggregation: How to aggregate results (simple, grouped, detailed)
|
101 |
+
min_score: Minimum confidence threshold (0.0-1.0)
|
102 |
+
|
103 |
+
Returns:
|
104 |
+
Formatted string with NER analysis results
|
105 |
+
"""
|
106 |
+
# Set default values if parameters are None
|
107 |
+
if model is None:
|
108 |
+
model = self.default_model
|
109 |
+
if aggregation is None:
|
110 |
+
aggregation = "grouped"
|
111 |
+
if min_score is None:
|
112 |
+
min_score = 0.8
|
113 |
+
|
114 |
+
# Validate model choice
|
115 |
+
if model not in self.available_models and not model.startswith("dslim/"):
|
116 |
+
return f"Model '{model}' not recognized. Available models: {', '.join(self.available_models.keys())}"
|
117 |
+
|
118 |
+
# Load the model if not already loaded or if different from current
|
119 |
+
if self._pipeline is None or self._pipeline.model.name_or_path != model:
|
120 |
+
if not self._load_pipeline(model):
|
121 |
+
return "Failed to load NER model. Please try a different model."
|
122 |
+
|
123 |
+
# Perform NER analysis
|
124 |
+
try:
|
125 |
+
entities = self._pipeline(text)
|
126 |
+
|
127 |
+
# Filter by confidence score
|
128 |
+
entities = [e for e in entities if e.get('score', 0) >= min_score]
|
129 |
+
|
130 |
+
if not entities:
|
131 |
+
return "No entities were detected in the text with the current settings."
|
132 |
+
|
133 |
+
# Process results based on aggregation method
|
134 |
+
if aggregation == "simple":
|
135 |
+
return self._format_simple(text, entities)
|
136 |
+
elif aggregation == "detailed":
|
137 |
+
return self._format_detailed(text, entities)
|
138 |
+
else: # default to grouped
|
139 |
+
return self._format_grouped(text, entities)
|
140 |
+
|
141 |
+
except Exception as e:
|
142 |
+
return f"Error analyzing text: {str(e)}"
|
143 |
+
|
144 |
+
def _format_simple(self, text: str, entities: List[Dict[str, Any]]) -> str:
|
145 |
+
"""Format entities as a simple list."""
|
146 |
+
result = "Named Entities Found:\n\n"
|
147 |
+
|
148 |
for entity in entities:
|
149 |
+
word = entity.get("word", "")
|
150 |
label = entity.get("entity", "UNKNOWN")
|
151 |
+
score = entity.get("score", 0)
|
152 |
+
friendly_label = self._get_friendly_label(label)
|
153 |
+
|
154 |
+
result += f"• {word} - {friendly_label} (confidence: {score:.2f})\n"
|
155 |
+
|
156 |
+
return result
|
157 |
+
|
158 |
+
def _format_grouped(self, text: str, entities: List[Dict[str, Any]]) -> str:
|
159 |
+
"""Format entities grouped by their category."""
|
160 |
+
# Group entities by their label
|
161 |
+
grouped = {}
|
162 |
+
|
163 |
+
for entity in entities:
|
164 |
word = entity.get("word", "")
|
165 |
+
label = entity.get("entity", "UNKNOWN").replace("B-", "").replace("I-", "")
|
166 |
+
|
167 |
+
if label not in grouped:
|
168 |
+
grouped[label] = []
|
169 |
+
|
170 |
+
grouped[label].append(word)
|
171 |
+
|
172 |
+
# Build the result string
|
173 |
+
result = "Named Entities by Category:\n\n"
|
174 |
+
|
175 |
+
for label, words in grouped.items():
|
176 |
+
friendly_label = self._get_friendly_label(label)
|
177 |
+
unique_words = list(set(words))
|
178 |
+
result += f"{friendly_label}: {', '.join(unique_words)}\n"
|
179 |
+
|
180 |
+
return result
|
181 |
+
|
182 |
+
def _format_detailed(self, text: str, entities: List[Dict[str, Any]]) -> str:
|
183 |
+
"""Format entities with detailed information including position in text."""
|
184 |
+
# First, build an entity map to highlight the entire text
|
185 |
+
character_labels = [None] * len(text)
|
186 |
+
|
187 |
+
# Mark each character with its entity
|
188 |
+
for entity in entities:
|
189 |
+
start = entity.get("start", 0)
|
190 |
+
end = entity.get("end", 0)
|
191 |
+
label = entity.get("entity", "UNKNOWN")
|
192 |
+
|
193 |
+
for i in range(start, min(end, len(text))):
|
194 |
+
character_labels[i] = label
|
195 |
+
|
196 |
+
# Build highlighted text sections
|
197 |
+
highlighted_text = ""
|
198 |
+
current_label = None
|
199 |
+
current_segment = ""
|
200 |
+
|
201 |
+
for i, char in enumerate(text):
|
202 |
+
label = character_labels[i]
|
203 |
+
|
204 |
+
if label != current_label:
|
205 |
+
# End the previous segment if any
|
206 |
+
if current_segment:
|
207 |
+
if current_label:
|
208 |
+
clean_label = current_label.replace("B-", "").replace("I-", "")
|
209 |
+
highlighted_text += f"[{current_segment}]({clean_label}) "
|
210 |
+
else:
|
211 |
+
highlighted_text += current_segment + " "
|
212 |
+
|
213 |
+
# Start a new segment
|
214 |
current_label = label
|
215 |
+
current_segment = char
|
216 |
else:
|
217 |
+
current_segment += char
|
218 |
+
|
219 |
+
# Add the final segment
|
220 |
+
if current_segment:
|
221 |
+
if current_label:
|
222 |
+
clean_label = current_label.replace("B-", "").replace("I-", "")
|
223 |
+
highlighted_text += f"[{current_segment}]({clean_label})"
|
224 |
+
else:
|
225 |
+
highlighted_text += current_segment
|
226 |
+
|
227 |
+
# Get entity details
|
228 |
+
entity_details = []
|
229 |
+
for entity in entities:
|
230 |
+
word = entity.get("word", "")
|
231 |
+
label = entity.get("entity", "UNKNOWN")
|
232 |
+
score = entity.get("score", 0)
|
233 |
+
friendly_label = self._get_friendly_label(label)
|
234 |
+
|
235 |
+
entity_details.append(f"• {word} - {friendly_label} (confidence: {score:.2f})")
|
236 |
+
|
237 |
+
# Combine into final result
|
238 |
+
result = "Entity Analysis:\n\n"
|
239 |
+
result += "Text with Entities Marked:\n"
|
240 |
+
result += highlighted_text + "\n\n"
|
241 |
+
result += "Entity Details:\n"
|
242 |
+
result += "\n".join(entity_details)
|
243 |
+
|
244 |
+
return result
|
245 |
+
|
246 |
+
def get_available_models(self) -> Dict[str, str]:
|
247 |
+
"""Return the dictionary of available models with descriptions."""
|
248 |
+
return self.available_models
|
249 |
|
250 |
+
# Example usage:
|
251 |
+
# ner_tool = NamedEntityRecognitionTool()
|
252 |
+
# result = ner_tool("Apple Inc. is planning to open a new store in Paris, France next year.", model="dslim/bert-base-NER")
|
253 |
+
# print(result)
|