Artifact cleanup

.semversioner/next-release/minor-20241113010525824646.json

@@ -0,0 +1,4 @@
+{
+  "type": "minor",
+  "description": "Data model changes."
+}

.semversioner/next-release/patch-20241031230557819462.json

@@ -0,0 +1,4 @@
+{
+  "type": "patch",
+  "description": "Cleanup of artifact outputs/schemas."
+}

docs/config/env_vars.md

@@ -9,8 +9,8 @@ If the embedding target is `all`, and you want to only embed a subset of these f
 ### Embedded Fields
 
 - `text_unit.text`
-- `document.raw_content`
-- `entity.name`
+- `document.text`
+- `entity.title`
 - `entity.description`
 - `relationship.description`
 - `community.title`

docs/examples_notebooks/drift_search.ipynb

@@ -204,7 +204,7 @@
     "# load description embeddings to an in-memory lancedb vectorstore\n",
     "# to connect to a remote db, specify url and port values.\n",
     "description_embedding_store = LanceDBVectorStore(\n",
-    "    collection_name=\"entity_description_embeddings\",\n",
+    "    collection_name=\"default-entity-description\",\n",
     ")\n",
     "description_embedding_store.connect(db_uri=LANCEDB_URI)\n",
     "entity_description_embeddings = store_entity_semantic_embeddings(\n",

docs/examples_notebooks/local_search.ipynb

@@ -108,7 +108,7 @@
     "# load description embeddings to an in-memory lancedb vectorstore\n",
     "# to connect to a remote db, specify url and port values.\n",
     "description_embedding_store = LanceDBVectorStore(\n",
-    "    collection_name=\"entity.description\",\n",
+    "    collection_name=\"default-entity-description\",\n",
     ")\n",
     "description_embedding_store.connect(db_uri=LANCEDB_URI)\n",
     "entity_description_embeddings = store_entity_semantic_embeddings(\n",

docs/index/default_dataflow.md

@@ -9,7 +9,8 @@ The knowledge model is a specification for data outputs that conform to our data
 - `Entity` - An entity extracted from a TextUnit. These represent people, places, events, or some other entity-model that you provide.
 - `Relationship` - A relationship between two entities. These are generated from the covariates.
 - `Covariate` - Extracted claim information, which contains statements about entities which may be time-bound.
-- `Community Report` - Once entities are generated, we perform hierarchical community detection on them and generate reports for each community in this hierarchy.
+- `Community` - Once the graph of entities and relationships is built, we perform hierarchical community detection on them to create a clustering structure.
+- `Community Report` - The contents of each community are summarized into a generated report, useful for human reading and downstream search.
 - `Node` - This table contains layout information for rendered graph-views of the Entities and Documents which have been embedded and clustered.
 
 ## The Default Configuration Workflow

docs/index/outputs.md

@@ -0,0 +1,89 @@
+# Outputs
+
+The default pipeline produces a series of output tables that align with the [conceptual knowledge model](../index/default_dataflow.md). This page describes the detailed output table schemas. By default we write these tables out as parquet files on disk.
+
+## Shared fields
+All tables have two identifier fields:
+- id: str - Generated UUID, assuring global uniqueness
+- human_readable_id: int - This is an incremented short ID created per-run. For example, we use this short ID with generated summaries that print citations so they are easy to cross-reference visually.
+
+## create_final_communities
+This is a list of the final communities generated by Leiden. Communities are strictly hierarchical, subdividing into children as the cluster affinity is narrowed.
+- community: int - Leiden-generated cluster ID for the community. Note that these increment with depth, so they are unique through all levels of the community hierarchy. For this table, human_readable_id is a copy of the community ID rather than a plain increment.
+- level: int - Depth of the community in the hierarchy.
+- title: str - Friendly name of the community.
+- entity_ids - List of entities that are members of the community.
+- relationship_ids - List of relationships that are wholly within the community (source and target are both in the community).
+- text_unit_ids - List of text units represented within the community.
+- period - Date of ingest, used for incremental update merges.
+- size  - Size of the community (entity count), used for incremental update merges.
+
+## create_final_community_reports
+This is the list of summarized reports for each community.
+- community: int - Short ID of the community this report applies to.
+- level: int - Level of the community this report applies to.
+- title: str - LM-generated title for the report.
+- summary: str - LM-generated summary of the report.
+- full_content: str - LM-generated full report.
+- rank: float - LM-derived relevance ranking of the report based on member entity salience
+- rank_explanation - LM-derived explanation of the rank.
+- findings: dict - LM-derived list of the top 5-10 insights from the community. Contains `summary` and `explanation` values.
+- full_content_json - Full JSON output as returned by the LM. Most fields are extracted into columns, but this JSON is sent for query summarization so we leave it to allow for prompt tuning to add fields/content by end users.
+- period - Date of ingest, used for incremental update merges.
+- size  - Size of the community (entity count), used for incremental update merges.
+
+## create_final_covariates
+(Optional) If claim extraction is turned on, this is a list of the extracted covariates. Note that claims are typically oriented around identifying malicious behavior such as fraud, so they are not useful for all datasets.
+- covariate_type: str - This is always "claim" with our default covariates.
+- type: str - Nature of the claim type.
+- description: str - LM-generated description of the behavior.
+- subject_id: str - Name of the source entity (that is performing the claimed behavior).
+- object_id: str - Name of the target entity (that the claimed behavior is performed on).
+- status: str [TRUE, FALSE, SUSPECTED] - LM-derived assessment of the correctness of the claim.
+- start_date: str (ISO8601) - LM-derived start of the claimed activity.
+- end_date: str (ISO8601) - LM-derived end of the claimed activity.
+- source_text: str - Short string of text containing the claimed behavior.
+- text_unit_id: str - ID of the text unit the claim text was extracted from.
+
+## create_final_documents
+List of document content after import.
+- title: str - Filename, unless otherwise configured during CSV import.
+- text: str - Full text of the document.
+- text_unit_ids: str[] - List of text units (chunks) that were parsed from the document.
+- attributes: dict (optional) - If specified during CSV import, this is a dict of attributes for the document.
+
+# create_final_entities
+List of all entities found in the data by the LM.
+- title: str - Name of the entity.
+- type: str - Type of the entity. By default this will be "organization", "person", "geo", or "event" unless configured differently or auto-tuning is used.
+- description: str - Textual description of the entity. Entities may be found in many text units, so this is an LM-derived summary of all descriptions.
+- text_unit_ids: str[] - List of the text units containing the entity.
+
+# create_final_nodes
+This is graph-related information for the entities. It contains only information relevant to the graph such as community. There is an entry for each entity at every community level it is found within, so you may see "duplicate" entities.
+
+Note that the ID fields match those in create_final_entities and can be used for joining if additional information about a node is required.
+- title: str - Name of the referenced entity. Duplicated from create_final_entities for convenient cross-referencing.
+- community: int - Leiden community the node is found within. Entities are not always assigned a community (they may not be close enough to any), so they may have a ID of -1.
+- level: int - Level of the community the entity is in.
+- degree: int - Node degree (connectedness) in the graph.
+- x: float - X position of the node for visual layouts. If graph embeddings and UMAP are not turned on, this will be 0.
+- y: float - Y position of the node for visual layouts. If graph embeddings and UMAP are not turned on, this will be 0.
+
+## create_final_relationships
+List of all entity-to-entity relationships found in the data by the LM. This is also the _edge list_ for the graph.
+- source: str - Name of the source entity.
+- target: str - Name of the target entity.
+- description: str - LM-derived description of the relationship. Also see note for entity descriptions.
+- weight: float - Weight of the edge in the graph. This is summed from an LM-derived "strength" measure for each relationship instance.
+- combined_degree: int - Sum of source and target node degrees.
+- text_unit_ids: str[] - List of text units the relationship was found within.
+
+## create_final_text_units
+List of all text chunks parsed from the input documents.
+- text: str - Raw full text of the chunk.
+- n_tokens: int - Number of tokens in the chunk. This should normally match the `chunk_size` config parameter, except for the last chunk which is often shorter.
+- document_ids: str[] - List of document IDs the chunk came from. This is normally only 1 due to our default groupby, but for very short text documents (e.g., microblogs) it can be configured so text units span multiple documents.
+- entity_ids: str[] - List of entities found in the text unit.
+- relationships_ids: str[] - List of relationships found in the text unit.
+- covariate_ids: str[] - Optional list of covariates found in the text unit.

examples_notebooks/community_contrib/yfiles-jupyter-graphs/graph-visualization.ipynb

@@ -299,7 +299,7 @@
     "entities = read_indexer_entities(entity_df, entity_embedding_df, COMMUNITY_LEVEL)\n",
     "\n",
     "description_embedding_store = LanceDBVectorStore(\n",
-    "    collection_name=\"entity.description\",\n",
+    "    collection_name=\"default-entity-description\",\n",
     ")\n",
     "description_embedding_store.connect(db_uri=LANCEDB_URI)\n",
     "entity_description_embeddings = store_entity_semantic_embeddings(\n",

graphrag/api/query.py

@@ -25,6 +25,10 @@
 from pydantic import validate_call
 
 from graphrag.config import GraphRagConfig
+from graphrag.index.config.embeddings import (
+    community_full_content_embedding,
+    entity_description_embedding,
+)
 from graphrag.logging import PrintProgressReporter
 from graphrag.query.factories import (
     get_drift_search_engine,
@@ -42,6 +46,7 @@
 )
 from graphrag.query.structured_search.base import SearchResult  # noqa: TCH001
 from graphrag.utils.cli import redact
+from graphrag.utils.embeddings import create_collection_name
 from graphrag.vector_stores import VectorStoreFactory, VectorStoreType
 from graphrag.vector_stores.base import BaseVectorStore
 
@@ -228,7 +233,7 @@ async def local_search(
 
     description_embedding_store = _get_embedding_store(
         config_args=vector_store_args,  # type: ignore
-        container_suffix="entity-description",
+        embedding_name=entity_description_embedding,
     )
 
     _entities = read_indexer_entities(nodes, entities, community_level)
@@ -302,7 +307,7 @@ async def local_search_streaming(
 
     description_embedding_store = _get_embedding_store(
         config_args=vector_store_args,  # type: ignore
-        container_suffix="entity-description",
+        embedding_name=entity_description_embedding,
     )
 
     _entities = read_indexer_entities(nodes, entities, community_level)
@@ -385,12 +390,12 @@ async def drift_search(
 
     description_embedding_store = _get_embedding_store(
         config_args=vector_store_args,  # type: ignore
-        container_suffix="entity-description",
+        embedding_name=entity_description_embedding,
     )
 
     full_content_embedding_store = _get_embedding_store(
         config_args=vector_store_args,  # type: ignore
-        container_suffix="community-full_content",
+        embedding_name=community_full_content_embedding,
     )
 
     _entities = read_indexer_entities(nodes, entities, community_level)
@@ -450,7 +455,10 @@ def _patch_vector_store(
         }
         description_embedding_store = LanceDBVectorStore(
             db_uri=config.embeddings.vector_store["db_uri"],
-            collection_name="default-entity-description",
+            collection_name=create_collection_name(
+                config.embeddings.vector_store["container_name"],
+                entity_description_embedding,
+            ),
             overwrite=config.embeddings.vector_store["overwrite"],
         )
         description_embedding_store.connect(
@@ -469,11 +477,7 @@ def _patch_vector_store(
             from graphrag.vector_stores.lancedb import LanceDBVectorStore
 
             community_reports = with_reports
-            collection_name = (
-                config.embeddings.vector_store.get("container_name", "default")
-                if config.embeddings.vector_store
-                else "default"
-            )
+            container_name = config.embeddings.vector_store["container_name"]
             # Store report embeddings
             _reports = read_indexer_reports(
                 community_reports,
@@ -485,7 +489,9 @@ def _patch_vector_store(
 
             full_content_embedding_store = LanceDBVectorStore(
                 db_uri=config.embeddings.vector_store["db_uri"],
-                collection_name=f"{collection_name}-community-full_content",
+                collection_name=create_collection_name(
+                    container_name, community_full_content_embedding
+                ),
                 overwrite=config.embeddings.vector_store["overwrite"],
             )
             full_content_embedding_store.connect(
@@ -501,12 +507,12 @@ def _patch_vector_store(
 
 def _get_embedding_store(
     config_args: dict,
-    container_suffix: str,
+    embedding_name: str,
 ) -> BaseVectorStore:
     """Get the embedding description store."""
     vector_store_type = config_args["type"]
-    collection_name = (
-        f"{config_args.get('container_name', 'default')}-{container_suffix}"
+    collection_name = create_collection_name(
+        config_args.get("container_name", "default"), embedding_name
     )
     embedding_store = VectorStoreFactory.get_vector_store(
         vector_store_type=vector_store_type,

graphrag/index/config/__init__.py

@@ -16,9 +16,9 @@
     community_full_content_embedding,
     community_summary_embedding,
     community_title_embedding,
-    document_raw_content_embedding,
+    document_text_embedding,
     entity_description_embedding,
-    entity_name_embedding,
+    entity_title_embedding,
     relationship_description_embedding,
     required_embeddings,
     text_unit_text_embedding,
@@ -82,9 +82,9 @@
     "community_full_content_embedding",
     "community_summary_embedding",
     "community_title_embedding",
-    "document_raw_content_embedding",
+    "document_text_embedding",
     "entity_description_embedding",
-    "entity_name_embedding",
+    "entity_title_embedding",
     "relationship_description_embedding",
     "required_embeddings",
     "text_unit_text_embedding",

graphrag/index/config/embeddings.py

@@ -3,20 +3,20 @@
 
 """A module containing embeddings values."""
 
-entity_name_embedding = "entity.name"
+entity_title_embedding = "entity.title"
 entity_description_embedding = "entity.description"
 relationship_description_embedding = "relationship.description"
-document_raw_content_embedding = "document.raw_content"
+document_text_embedding = "document.text"
 community_title_embedding = "community.title"
 community_summary_embedding = "community.summary"
 community_full_content_embedding = "community.full_content"
 text_unit_text_embedding = "text_unit.text"
 
 all_embeddings: set[str] = {
-    entity_name_embedding,
+    entity_title_embedding,
     entity_description_embedding,
     relationship_description_embedding,
-    document_raw_content_embedding,
+    document_text_embedding,
     community_title_embedding,
     community_summary_embedding,
     community_full_content_embedding,

graphrag/index/flows/create_base_entity_graph.py

@@ -30,8 +30,6 @@ async def create_base_entity_graph(
     callbacks: VerbCallbacks,
     cache: PipelineCache,
     storage: PipelineStorage,
-    text_column: str,
-    id_column: str,
     clustering_strategy: dict[str, Any],
     extraction_strategy: dict[str, Any] | None = None,
     extraction_num_threads: int = 4,
@@ -52,8 +50,8 @@ async def create_base_entity_graph(
         text_units,
         callbacks,
         cache,
-        text_column=text_column,
-        id_column=id_column,
+        text_column="text",
+        id_column="id",
         strategy=extraction_strategy,
         async_mode=extraction_async_mode,
         entity_types=entity_types,

graphrag/index/flows/create_base_text_units.py

@@ -24,8 +24,6 @@ async def create_base_text_units(
     documents: pd.DataFrame,
     callbacks: VerbCallbacks,
     storage: PipelineStorage,
-    chunk_column_name: str,
-    n_tokens_column_name: str,
     chunk_by_columns: list[str],
     chunk_strategy: dict[str, Any] | None = None,
     snapshot_transient_enabled: bool = False,
@@ -65,21 +63,18 @@ async def create_base_text_units(
     chunked = chunked.explode("chunks")
     chunked.rename(
         columns={
-            "chunks": chunk_column_name,
+            "chunks": "chunk",
         },
         inplace=True,
     )
-    chunked["chunk_id"] = chunked.apply(
-        lambda row: gen_md5_hash(row, [chunk_column_name]), axis=1
+    chunked["id"] = chunked.apply(lambda row: gen_md5_hash(row, ["chunk"]), axis=1)
+    chunked[["document_ids", "chunk", "n_tokens"]] = pd.DataFrame(
+        chunked["chunk"].tolist(), index=chunked.index
     )
-    chunked[["document_ids", chunk_column_name, n_tokens_column_name]] = pd.DataFrame(
-        chunked[chunk_column_name].tolist(), index=chunked.index
-    )
-    chunked["id"] = chunked["chunk_id"]
+    # rename for downstream consumption
+    chunked.rename(columns={"chunk": "text"}, inplace=True)
 
-    output = cast(
-        pd.DataFrame, chunked[chunked[chunk_column_name].notna()].reset_index(drop=True)
-    )
+    output = cast(pd.DataFrame, chunked[chunked["text"].notna()].reset_index(drop=True))
 
     if snapshot_transient_enabled:
         await snapshot(