allow to add vector embeddings and metadata without content in vector databases.

Open
#2,555 0 comments 1 reaction 0 assignees View on GitHub

Nobody has claimed this yet.

Assessment

Difficulty
4/5
Estimated time
3-5 days
Newbie friendliness
35/100
Issue type
Feature
Clarity
Mostly clear
Activity status
Stale
Tech stack
java, spring
Domain
ai, backend-api-design, databases

Research direction

Start by reading the VectorStore API and the Weaviate vector store integration mentioned in the issue, then trace the current add(documents) flow. Done should allow embeddings and metadata to be stored without duplicated content, while preserving empty content where a vector database requires it.

Written by the indexing model from the issue text.

Description

status: waiting-for-triage

Expected Behavior

VectorStore api should have a flag that will allow to store vector embeddings without content(or empty content in case vector database doesn't support embeddings store without content).

for vector store, we can have a bool flag that will allow to sotore embeddings and metadata without content(by setting content as empty internally)

bool storeContent = false;
vectorStore.add(documents,storeContent);

Current Behavior

the given current implementation of vector store to add data is:

VectorStore vectorStore;
List<Doucment> documents = List.of(d1,d2);
vectorStore.add(documents);

Context

I am building an application to store religious scriptures along with original languages as well as their translations with contextual search features.
Given the complex nature of data relations, i am storing all data into RDBMS as it makes data management easy.
Given that metadata allows us to uniquely identify and fetch the content from vector database, i am storing vector embeddings along with full metadata into vector database via spring ai api.
However, current implementation also stores content along with metadata and embeddings, which results in huge data duplication, which i am trying to avoid.
I considered manually generating emebddings and store them into vector database without content, but current api implementation is too complex and documentation lacks any example/guideline to help implement that custom implementation.
No, i am not aware of any workarounds at the moment.

Note: I am using weaviate vector store api, but it should be true for all vector database embeddings.

Thank you

Dominant language
Java
Stars
9.5k
Forks
2.9k
Avg merge
1d 10h
Merged PRs (30d)
5

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

More from spring-projects/spring-ai

All issues in spring-projects/spring-ai

Similar issues

More Java issues

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.