Models: BM25 on segmented documents with doc2query-T5 expansions
This page describes regression experiments for document ranking on the segmented version of the MS MARCO (V2) document corpus using the dev queries, which is integrated into Anserini's regression testing framework. Here, we expand the segmented document corpus with doc2query-T5.
The exact configurations for these regressions are stored in this YAML file. Note that this page is automatically generated from this template as part of Anserini's regression pipeline, so do not modify this page directly; modify the template instead.
From one of our Waterloo servers (e.g., orca), the following command will perform the complete regression, end to end:
python src/main/python/run_regression.py --index --verify --search --regression msmarco-v2-doc-segmented-d2q-t5
Typical indexing command:
target/appassembler/bin/IndexCollection \
-collection MsMarcoV2DocCollection \
-input /path/to/msmarco-v2-doc-segmented-d2q-t5 \
-index indexes/lucene-index.msmarco-v2-doc-segmented-d2q-t5/ \
-generator DefaultLuceneDocumentGenerator \
-threads 24 -storeRaw \
>& logs/log.msmarco-v2-doc-segmented-d2q-t5 &
The directory /path/to/msmarco-v2-doc-segmented-d2q-t5/ should be a directory containing the compressed jsonl files that comprise the corpus.
See this page for additional details.
For additional details, see explanation of common indexing options.
Topics and qrels are stored here, which is linked to the Anserini repo as a submodule.
After indexing has completed, you should be able to perform retrieval as follows:
target/appassembler/bin/SearchCollection \
-index indexes/lucene-index.msmarco-v2-doc-segmented-d2q-t5/ \
-topics tools/topics-and-qrels/topics.msmarco-v2-doc.dev.txt \
-topicreader TsvInt \
-output runs/run.msmarco-v2-doc-segmented-d2q-t5.bm25-default.topics.msmarco-v2-doc.dev.txt \
-bm25 -hits 10000 -selectMaxPassage -selectMaxPassage.delimiter "#" -selectMaxPassage.hits 1000 &
target/appassembler/bin/SearchCollection \
-index indexes/lucene-index.msmarco-v2-doc-segmented-d2q-t5/ \
-topics tools/topics-and-qrels/topics.msmarco-v2-doc.dev2.txt \
-topicreader TsvInt \
-output runs/run.msmarco-v2-doc-segmented-d2q-t5.bm25-default.topics.msmarco-v2-doc.dev2.txt \
-bm25 -hits 10000 -selectMaxPassage -selectMaxPassage.delimiter "#" -selectMaxPassage.hits 1000 &
Evaluation can be performed using trec_eval:
tools/eval/trec_eval.9.0.4/trec_eval -c -m recall.100 tools/topics-and-qrels/qrels.msmarco-v2-doc.dev.txt runs/run.msmarco-v2-doc-segmented-d2q-t5.bm25-default.topics.msmarco-v2-doc.dev.txt
tools/eval/trec_eval.9.0.4/trec_eval -c -m recall.1000 tools/topics-and-qrels/qrels.msmarco-v2-doc.dev.txt runs/run.msmarco-v2-doc-segmented-d2q-t5.bm25-default.topics.msmarco-v2-doc.dev.txt
tools/eval/trec_eval.9.0.4/trec_eval -c -M 100 -m map -c -M 100 -m recip_rank tools/topics-and-qrels/qrels.msmarco-v2-doc.dev.txt runs/run.msmarco-v2-doc-segmented-d2q-t5.bm25-default.topics.msmarco-v2-doc.dev.txt
tools/eval/trec_eval.9.0.4/trec_eval -c -m recall.100 tools/topics-and-qrels/qrels.msmarco-v2-doc.dev2.txt runs/run.msmarco-v2-doc-segmented-d2q-t5.bm25-default.topics.msmarco-v2-doc.dev2.txt
tools/eval/trec_eval.9.0.4/trec_eval -c -m recall.1000 tools/topics-and-qrels/qrels.msmarco-v2-doc.dev2.txt runs/run.msmarco-v2-doc-segmented-d2q-t5.bm25-default.topics.msmarco-v2-doc.dev2.txt
tools/eval/trec_eval.9.0.4/trec_eval -c -M 100 -m map -c -M 100 -m recip_rank tools/topics-and-qrels/qrels.msmarco-v2-doc.dev2.txt runs/run.msmarco-v2-doc-segmented-d2q-t5.bm25-default.topics.msmarco-v2-doc.dev2.txt
With the above commands, you should be able to reproduce the following results:
| MAP@100 | BM25 (default) |
|---|---|
| MS MARCO V2 Doc: Dev | 0.2203 |
| MS MARCO V2 Doc: Dev2 | 0.2205 |
| MRR@100 | BM25 (default) |
| MS MARCO V2 Doc: Dev | 0.2226 |
| MS MARCO V2 Doc: Dev2 | 0.2234 |
| R@100 | BM25 (default) |
| MS MARCO V2 Doc: Dev | 0.7297 |
| MS MARCO V2 Doc: Dev2 | 0.7316 |
| R@1000 | BM25 (default) |
| MS MARCO V2 Doc: Dev | 0.8982 |
| MS MARCO V2 Doc: Dev2 | 0.8952 |