JingsongLi opened a new pull request, #8351:
URL: https://github.com/apache/paimon/pull/8351

   ## Summary
   Adds a pluggable hybrid search reranker interface and extends built-in 
reranking with LanceDB-compatible score/rank fusion options. This lets hybrid 
search use RRF, weighted score/linear combination, MRR, or a custom Java 
reranker before returning the final top-k rows.
   
   ## Changes
   - Added `HybridSearchReranker` and routed hybrid result fusion through the 
reranker abstraction.
   - Added built-in `mrr` reranking plus `linear_combination` as an alias for 
`weighted_score`.
   - Exposed `withReranker(...)` and `withMrrRanker()` on `HybridSearchBuilder`.
   - Updated Spark TVF parsing and hybrid search docs for the new reranker 
names and Java custom reranker entry point.
   
   ## Testing
   - `mvn -pl paimon-common -Pfast-build -Dtest=HybridSearchRankerTest test`
   - `mvn -pl paimon-core -am -Pfast-build -DfailIfNoTests=false 
-Dtest=VectorSearchBuilderTest#testHybridSearchBuilderExposesRouteBuilders test`
   - `mvn -pl paimon-spark/paimon-spark-common -am -Pfast-build 
-DfailIfNoTests=false 
-DwildcardSuites=org.apache.paimon.spark.catalyst.plans.logical.VectorSearchQueryTest
 -Dtest=none test`
   - `git diff --check`
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to