File size: 7,363 Bytes
37f0321
 
 
 
d02622b
37f0321
 
 
 
 
 
 
 
 
 
 
 
d02622b
8c66169
 
37f0321
 
 
 
 
 
 
 
 
 
 
 
 
0b9d8c7
37f0321
 
 
 
 
 
 
0b9d8c7
37f0321
 
 
 
 
 
 
 
 
 
 
 
 
 
0b9d8c7
37f0321
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
9e67c3b
37f0321
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
9e67c3b
37f0321
 
 
 
 
 
 
 
 
 
 
9e67c3b
37f0321
 
 
 
 
 
 
 
 
 
 
 
 
 
9e67c3b
 
37f0321
 
 
 
 
 
 
 
 
 
 
 
9e67c3b
37f0321
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
import re
from collections import Counter, defaultdict
from typing import Dict, List, Tuple

from graphgen.bases import BaseGraphStorage, BaseKGBuilder, BaseLLMWrapper, Chunk
from graphgen.templates import KG_EXTRACTION_PROMPT, KG_SUMMARIZATION_PROMPT
from graphgen.utils import (
    detect_main_language,
    handle_single_entity_extraction,
    handle_single_relationship_extraction,
    logger,
    pack_history_conversations,
    split_string_by_multi_markers,
)


class LightRAGKGBuilder(BaseKGBuilder):
    def __init__(self, llm_client: BaseLLMWrapper, max_loop: int = 3):
        super().__init__(llm_client)
        self.max_loop = max_loop

    async def extract(
        self, chunk: Chunk
    ) -> Tuple[Dict[str, List[dict]], Dict[Tuple[str, str], List[dict]]]:
        """
        Extract entities and relationships from a single chunk using the LLM client.
        :param chunk
        :return: (nodes_data, edges_data)
        """
        chunk_id = chunk.id
        content = chunk.content

        # step 1: language_detection
        language = detect_main_language(content)

        hint_prompt = KG_EXTRACTION_PROMPT[language]["TEMPLATE"].format(
            **KG_EXTRACTION_PROMPT["FORMAT"], input_text=content
        )

        # step 2: initial glean
        final_result = await self.llm_client.generate_answer(hint_prompt)
        logger.debug("First extraction result: %s", final_result)

        # step3: iterative refinement
        history = pack_history_conversations(hint_prompt, final_result)
        for loop_idx in range(self.max_loop):
            if_loop_result = await self.llm_client.generate_answer(
                text=KG_EXTRACTION_PROMPT[language]["IF_LOOP"], history=history
            )
            if_loop_result = if_loop_result.strip().strip('"').strip("'").lower()
            if if_loop_result != "yes":
                break

            glean_result = await self.llm_client.generate_answer(
                text=KG_EXTRACTION_PROMPT[language]["CONTINUE"], history=history
            )
            logger.debug("Loop %s glean: %s", loop_idx + 1, glean_result)

            history += pack_history_conversations(
                KG_EXTRACTION_PROMPT[language]["CONTINUE"], glean_result
            )
            final_result += glean_result

        # step 4: parse the final result
        records = split_string_by_multi_markers(
            final_result,
            [
                KG_EXTRACTION_PROMPT["FORMAT"]["record_delimiter"],
                KG_EXTRACTION_PROMPT["FORMAT"]["completion_delimiter"],
            ],
        )

        nodes = defaultdict(list)
        edges = defaultdict(list)

        for record in records:
            match = re.search(r"\((.*)\)", record)
            if not match:
                continue
            inner = match.group(1)

            attributes = split_string_by_multi_markers(
                inner, [KG_EXTRACTION_PROMPT["FORMAT"]["tuple_delimiter"]]
            )

            entity = await handle_single_entity_extraction(attributes, chunk_id)
            if entity is not None:
                nodes[entity["entity_name"]].append(entity)
                continue

            relation = await handle_single_relationship_extraction(attributes, chunk_id)
            if relation is not None:
                key = (relation["src_id"], relation["tgt_id"])
                edges[key].append(relation)

        return dict(nodes), dict(edges)

    async def merge_nodes(
        self,
        node_data: tuple[str, List[dict]],
        kg_instance: BaseGraphStorage,
    ) -> None:
        entity_name, node_data = node_data
        entity_types = []
        source_ids = []
        descriptions = []

        node = kg_instance.get_node(entity_name)
        if node is not None:
            entity_types.append(node["entity_type"])
            source_ids.extend(
                split_string_by_multi_markers(node["source_id"], ["<SEP>"])
            )
            descriptions.append(node["description"])

        # take the most frequent entity_type
        entity_type = sorted(
            Counter([dp["entity_type"] for dp in node_data] + entity_types).items(),
            key=lambda x: x[1],
            reverse=True,
        )[0][0]

        description = "<SEP>".join(
            sorted(set([dp["description"] for dp in node_data] + descriptions))
        )
        description = await self._handle_kg_summary(entity_name, description)

        source_id = "<SEP>".join(
            set([dp["source_id"] for dp in node_data] + source_ids)
        )

        node_data = {
            "entity_type": entity_type,
            "description": description,
            "source_id": source_id,
        }
        kg_instance.upsert_node(entity_name, node_data=node_data)

    async def merge_edges(
        self,
        edges_data: tuple[Tuple[str, str], List[dict]],
        kg_instance: BaseGraphStorage,
    ) -> None:
        (src_id, tgt_id), edge_data = edges_data

        source_ids = []
        descriptions = []

        edge = kg_instance.get_edge(src_id, tgt_id)
        if edge is not None:
            source_ids.extend(
                split_string_by_multi_markers(edge["source_id"], ["<SEP>"])
            )
            descriptions.append(edge["description"])

        description = "<SEP>".join(
            sorted(set([dp["description"] for dp in edge_data] + descriptions))
        )
        source_id = "<SEP>".join(
            set([dp["source_id"] for dp in edge_data] + source_ids)
        )

        for insert_id in [src_id, tgt_id]:
            if not kg_instance.has_node(insert_id):
                kg_instance.upsert_node(
                    insert_id,
                    node_data={
                        "source_id": source_id,
                        "description": description,
                        "entity_type": "UNKNOWN",
                    },
                )

        description = await self._handle_kg_summary(
            f"({src_id}, {tgt_id})", description
        )

        kg_instance.upsert_edge(
            src_id,
            tgt_id,
            edge_data={"source_id": source_id, "description": description},
        )

    async def _handle_kg_summary(
        self,
        entity_or_relation_name: str,
        description: str,
        max_summary_tokens: int = 200,
    ) -> str:
        """
        Handle knowledge graph summary

        :param entity_or_relation_name
        :param description
        :param max_summary_tokens
        :return summary
        """

        tokenizer_instance = self.llm_client.tokenizer
        language = detect_main_language(description)

        tokens = tokenizer_instance.encode(description)
        if len(tokens) < max_summary_tokens:
            return description

        use_description = tokenizer_instance.decode(tokens[:max_summary_tokens])
        prompt = KG_SUMMARIZATION_PROMPT[language]["TEMPLATE"].format(
            entity_name=entity_or_relation_name,
            description_list=use_description.split("<SEP>"),
            **KG_SUMMARIZATION_PROMPT["FORMAT"],
        )
        new_description = await self.llm_client.generate_answer(prompt)
        logger.info(
            "Entity or relation %s summary: %s",
            entity_or_relation_name,
            new_description,
        )
        return new_description