Update README.md

README.md

@@ -12,100 +12,40 @@ size_categories:
   - 100M<n<1B
 ---
 
-# Dataset Card for StackOverflow Knowledge Graph
+# Dataset Card for StackOverflow Vector Snapshot Dataset
 
-## Dataset Summary
+## Summary
 
-This dataset contains the knowledge graph component of Stack2Graph, a structured representation of Stack Overflow data introduced in the paper Stack2Graph: A Structured Knowledge Representation of Stack Overflow Data for Retrieval-based Question Answering.
+This dataset contains the vector-database component of Stack2Graph as restorable Qdrant snapshots.
 
-The graph is generated from the Stack Overflow data dump after SQL ingestion and serialized as TriG RDF datasets. It preserves structural relations between questions, answers, comments, tags, votes, duplicates, and question-to-question links instead of reducing Stack Overflow to isolated question-answer pairs.
+The artifacts are intended for direct restore into Qdrant rather than row-wise reuse as a plain text or parquet dataset.
 
-As described in the paper, Stack2Graph combines a knowledge graph with a vector database. This Hugging Face dataset card covers only the knowledge graph artifact. The paper reports a resource scale of more than 117 million nodes and 275 million edges for Stack2Graph.
+The repository covers the vector side of Stack2Graph only, not the accompanying knowledge graph artifact.
 
 ## Repository Layout
 
-The uploaded repository stores compressed TriG shards under the `trig_files` directory.
-
 ```text
-trig_files/
+qdrant_snapshots/
+  collections_manifest.json
   archive_manifest.json
-  part_0.tar.gz
-  part_1.tar.gz
-  part_2.tar.gz
+  stackoverflow_python_vector.tar.xz
+  stackoverflow_java_vector.tar.xz
+  stackoverflow_javascript_vector.snapshot.xz.part_000
   ...
 ```
 
-Each archive contains multiple TriG shard files such as `part0.trig`, `part1.trig`, and `part2.trig`.
-
-The `archive_manifest.json` file records:
-
-- the archive batch size
-- the number of archive files
-- the number of source `.trig` files
-- the exact shard files contained in each archive
-
-## Graph Organization
-
-Each `.trig` file is a serialized RDF dataset containing:
-
-- a default graph with schema definitions
-- one named graph per supported programming-language tag
-
-Named graph IRIs follow this pattern:
-
-```text
-http://stackoverflow.com/<language>
-```
-
-Examples:
-
-- `http://stackoverflow.com/python`
-- `http://stackoverflow.com/java`
-- `http://stackoverflow.com/javascript`
-
-A question can appear in more than one named graph when it carries multiple supported language tags.
-
-## Graph Content
-
-For each retained question, the graph may contain:
-
-- the question node
-- accepted and non-accepted answers
-- comments on the question
-- comments on the answers
-- attached tags
-- tag wiki content for supported programming-language tags when available
-- aggregate vote counts
-- related-question links
-- duplicate-question links
-
-Questions are only included if they contain at least one supported programming-language tag.
-
-## Supported Language Tags
-
-The current pipeline retains questions when at least one tag matches this set:
-
-- `java`, `c`, `c++`, `python`, `c#`, `javascript`, `assembly`, `php`, `perl`, `ruby`, `vb.net`, `swift`, `r`, `objective-c`, `go`, `sql`, `matlab`, `typescript`, `scala`, `kotlin`, `rust`, `lua`, `haskell`, `cobol`, `fortran`, `lisp`, `erlang`, `elixir`, `f#`, `dart`, `shell`, `bash`, `powershell`, `css`, `html`, `.net`, `julia`, `prolog`, `abap`
+Each language is stored as exactly one restore unit:
 
-## Ontology and Namespaces
+- a single `tar.xz` archive for regular-size snapshots
+- split `xz` parts for oversized snapshots
 
-The dataset uses a Stack Overflow-specific namespace and resource layout.
+## Content
 
-Core namespaces:
+The snapshots represent per-language Qdrant collections derived from Stack Overflow questions tagged with supported programming-language tags.
 
-- `http://stackoverflow.com/schema#` as the ontology namespace (`so:`)
-- `http://stackoverflow.com/q/` for question instances
-- `http://stackoverflow.com/a/` for answer instances
-- `http://stackoverflow.com/comment/` for comment instances
-- `http://stackoverflow.com/tag/` for tag instances
+Each collection stores the vector index used by the Stack2Graph retrieval pipeline.
 
-Example resource IRIs:
-
-- Question: `http://stackoverflow.com/q/12345`
-- Answer: `http://stackoverflow.com/a/67890`
-- Comment: `http://stackoverflow.com/comment/111`
-- Tag: `http://stackoverflow.com/tag/python`
-- Tag wiki: `http://stackoverflow.com/tag/python/info`
+The current pipeline covers major programming-language tags such as Python, Java, JavaScript, C, C++, C#, Go, Rust, TypeScript, SQL, Shell, and additional languages used throughout the project.
 
 ## Schema Overview
 
@@ -113,95 +53,24 @@ Example resource IRIs:
 
 *Figure 1. High-level schema of Stack2Graph. Questions are the central entities and connect to answers, comments, tags, tag-wiki content, vote aggregates, and question-to-question links. Schema triples are stored in the default graph, while content triples are written into one named graph per supported programming-language tag.*
 
-Core entity types:
-
-- `so:Question`
-- `so:Answer`
-- `so:Comment`
-- `so:Tag`
-- `so:TagWiki`
-
-Key structural relations:
-
-- question-answer: `so:acceptedAnswer`, `so:hasAnswer`, `so:answersQuestion`, `so:possibleBetterAnswer`
-- post-comment: `so:hasComment`, `so:commentOn`
-- question-tag: `so:taggedWith`, `so:tagsQuestion`
-- tag-wiki: `so:hasTagWiki`, `so:tagWikiOf`
-- question-question: `so:linkedTo`, `so:duplicateOf`
-
-Core literal properties:
-
-- textual content: `so:hasTitle`, `so:hasBody`, `so:hasText`
-- interaction signals: `so:score`, `so:answerCount`, `so:upvoteCount`, `so:downvoteCount`, `so:favoriteCount`
-- tag label: `so:tagName`
-
-Additional ontology notes:
-
-- `so:hasAnswer` is the inverse of `so:answersQuestion`
-- `so:hasComment` is the inverse of `so:commentOn`
-- `so:hasTagWiki` is the inverse of `so:tagWikiOf`
-- `so:taggedWith` is the inverse of `so:tagsQuestion`
-- `so:linkedTo` is modeled as a symmetric property
-
-## Construction Rules
-
-The current generation pipeline applies the following rules:
-
-1. Questions are loaded from SQL in chunks.
-2. Only questions with at least one recognized programming-language tag are retained.
-3. Related answers, comments, links, and vote aggregates are fetched in batches.
-4. Tag resources are created for all question tags in the retained set.
-5. Tag wiki nodes are added for supported programming-language tags when a tag wiki post exists.
-6. Each retained question is written into one named graph per matching programming-language tag.
-7. Schema triples are included in the default graph of each TriG shard.
-
-Special cases:
-
-- `so:possibleBetterAnswer` is a heuristic edge used when a question has an accepted answer and another answer with a higher score.
-- Stack Overflow post link type `1` becomes `so:linkedTo`.
-- Stack Overflow post link type `3` becomes `so:duplicateOf`.
-
-## Data Generation
-
-The graph is generated from the Stack Overflow XML dump after ingestion into a SQL database.
-
-Primary source tables include Stack Overflow post, comment, vote, tag, and post-link data.
-
-The dataset is produced by these repository components:
-
-- `knowledgeGraph/xmlToSql.py`
-- `knowledgeGraph/generateKnowledgeGraph.py`
-- `knowledgeGraph/graphHelper.py`
-- `knowledgeGraph/upload_hf_trig_files.py`
-
-Serialization and packaging:
 
-- RDF format: TriG
-- repository packaging: `.tar.gz`
-- sharding: multiple `.trig` shards grouped into archive parts
+## Intended Use
 
-## Limitations
+This dataset is meant to be downloaded and restored into a running Qdrant instance.
 
-- The graph only includes questions that match the repository's supported language-tag list.
-- A single question may be duplicated across multiple named graphs when it has multiple language tags.
-- Bodies and comment text are stored as escaped HTML, not plain text.
-- The graph uses aggregate vote counts instead of individual vote events.
-- `so:possibleBetterAnswer` is heuristic and should not be treated as ground truth.
-- Tag wiki coverage depends on whether a corresponding wiki post exists in the source dump.
-- The dataset inherits the biases, moderation artifacts, and temporal drift of Stack Overflow content.
+The repository manifest files describe which artifact files belong to which Qdrant collection.
 
-## Figures
+It is primarily intended for retrieval experiments and system reconstruction, not for direct training from the packaged artifacts.
 
-This card is intentionally concise. Figures exported from the paper can be embedded here later to replace part of the textual explanation of the graph structure and pipeline.
+## Source
 
-## Licensing
+The snapshots are generated from Stack Overflow data after preprocessing, embedding, and Qdrant ingestion in the Stack2Graph pipeline.
 
-This dataset is distributed under the `CC-BY-SA-4.0` license.
+## License
 
-If you redistribute derived artifacts or use this dataset in downstream resources, you should preserve the attribution and share-alike requirements of `CC-BY-SA-4.0`.
+This dataset is distributed under `CC-BY-SA-4.0`.
 
 ## Citation
 
-If you use this dataset, cite:
+If you use this dataset, cite the Stack2Graph project and preserve the attribution requirements of `CC-BY-SA-4.0`.
 
-TODO