Skip to content

Latest commit

 

History

History
85 lines (66 loc) · 5.4 KB

regressions-msmarco-v2-passage.md

File metadata and controls

85 lines (66 loc) · 5.4 KB
Raw

Anserini Regressions: MS MARCO (V2) Passage Ranking

Models: various bag-of-words approaches on original passages

This page describes regression experiments for passage ranking on the MS MARCO V2 Passage Corpus using the dev queries, which is integrated into Anserini's regression testing framework. Here, we cover bag-of-words baselines. For more complete instructions on how to run end-to-end experiments, refer to this page.

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-passage

Indexing

Typical indexing command:

target/appassembler/bin/IndexCollection \
  -collection MsMarcoV2PassageCollection \
  -input /path/to/msmarco-v2-passage \
  -index indexes/lucene-index.msmarco-v2-passage/ \
  -generator DefaultLuceneDocumentGenerator \
  -threads 24 -storeRaw \
  >& logs/log.msmarco-v2-passage &

The directory /path/to/msmarco-v2-passage/ 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.

Retrieval

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-passage/ \
  -topics tools/topics-and-qrels/topics.msmarco-v2-passage.dev.txt \
  -topicreader TsvInt \
  -output runs/run.msmarco-v2-passage.bm25-default.topics.msmarco-v2-passage.dev.txt \
  -bm25 &
target/appassembler/bin/SearchCollection \
  -index indexes/lucene-index.msmarco-v2-passage/ \
  -topics tools/topics-and-qrels/topics.msmarco-v2-passage.dev2.txt \
  -topicreader TsvInt \
  -output runs/run.msmarco-v2-passage.bm25-default.topics.msmarco-v2-passage.dev2.txt \
  -bm25 &

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-passage.dev.txt runs/run.msmarco-v2-passage.bm25-default.topics.msmarco-v2-passage.dev.txt
tools/eval/trec_eval.9.0.4/trec_eval -c -m recall.1000 tools/topics-and-qrels/qrels.msmarco-v2-passage.dev.txt runs/run.msmarco-v2-passage.bm25-default.topics.msmarco-v2-passage.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-passage.dev.txt runs/run.msmarco-v2-passage.bm25-default.topics.msmarco-v2-passage.dev.txt
tools/eval/trec_eval.9.0.4/trec_eval -c -m recall.100 tools/topics-and-qrels/qrels.msmarco-v2-passage.dev2.txt runs/run.msmarco-v2-passage.bm25-default.topics.msmarco-v2-passage.dev2.txt
tools/eval/trec_eval.9.0.4/trec_eval -c -m recall.1000 tools/topics-and-qrels/qrels.msmarco-v2-passage.dev2.txt runs/run.msmarco-v2-passage.bm25-default.topics.msmarco-v2-passage.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-passage.dev2.txt runs/run.msmarco-v2-passage.bm25-default.topics.msmarco-v2-passage.dev2.txt

Effectiveness

With the above commands, you should be able to reproduce the following results:

MAP@100 BM25 (default)
MS MARCO V2 Passage: Dev 0.0709
MS MARCO V2 Passage: Dev2 0.0794
MRR@100 BM25 (default)
MS MARCO V2 Passage: Dev 0.0719
MS MARCO V2 Passage: Dev2 0.0802
R@100 BM25 (default)
MS MARCO V2 Passage: Dev 0.3397
MS MARCO V2 Passage: Dev2 0.3459
R@1000 BM25 (default)
MS MARCO V2 Passage: Dev 0.5733
MS MARCO V2 Passage: Dev2 0.5839