bm25: Switch from subword tokenizer to IntlSegmenter - #12
Merged
Conversation
Replace BPE subword tokenization with Intl.Segmenter for BM25 indexing. Subword tokenization breaks words into fragments (e.g., "frequency" -> "fre|que|ncy"), which hurts BM25 word matching. IntlSegmenter uses ICU word boundary detection, preserving whole words for better retrieval. Changes: - Add IntlSegmenterTokenizer using built-in Intl.Segmenter API - Update BM25Store to use IntlSegmenterTokenizer instead of Embedder - Remove Embedder dependency from BM25Store - Add rebuildBM25Index method to IndexManager Benchmark result (MIRACL Merged, nDCG@10): - Before (BGE-M3 subword): 0.8827 - After (IntlSegmenter): 0.9023 (+2.2%)
aviatesk
force-pushed
the
avi/bm25-tokenizer
branch
from
December 11, 2025 09:37
f8c7776 to
5026863
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Replace BPE subword tokenization with Intl.Segmenter for BM25 indexing. Subword tokenization breaks words into fragments (e.g., "frequency" -> "fre|que|ncy"), which hurts BM25 word matching. IntlSegmenter uses ICU word boundary detection, preserving whole words for better retrieval.
Changes:
Benchmark result (MIRACL Merged, nDCG@10):