Skip to content

Latest commit

 

History

History
177 lines (149 loc) · 13.2 KB

regressions-neuclir22-zh-dt.md

File metadata and controls

177 lines (149 loc) · 13.2 KB
Raw

Anserini Regressions: NeuCLIR22 — Chinese (Document Translation)

This page presents document translation regression experiments for the TREC 2022 NeuCLIR Track, Chinese, with the following configuration:

  • Queries: English
  • Documents: Machine-translated documents from Chinese into English (corpus provided by the organizers)
  • Model: BM25

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 neuclir22-zh-dt

Corpus Download

The NeuCLIR 2022 corpus can be downloaded following the instructions here.

With the corpus downloaded, unpack into collections/ and run the following command to perform the remaining steps below:

python src/main/python/run_regression.py --index --verify --search --regression neuclir22-zh-dt \
  --corpus-path collections/neuclir22-zh-en

Indexing

Typical indexing command:

bin/run.sh io.anserini.index.IndexCollection \
  -threads 8 \
  -collection NeuClirCollection \
  -input /path/to/neuclir22-zh-en \
  -generator DefaultLuceneDocumentGenerator \
  -index indexes/lucene-index.neuclir22-zh-en \
  -storePositions -storeDocvectors -storeRaw \
  >& logs/log.neuclir22-zh-en &

For additional details, see explanation of common indexing options.

Retrieval

After indexing has completed, you should be able to perform retrieval as follows:

bin/run.sh io.anserini.search.SearchCollection \
  -index indexes/lucene-index.neuclir22-zh-en \
  -topics tools/topics-and-qrels/topics.neuclir22-en.original-title.txt \
  -topicReader TsvInt \
  -output runs/run.neuclir22-zh-en.bm25-default.topics.neuclir22-en.original-title.txt \
  -bm25 &
bin/run.sh io.anserini.search.SearchCollection \
  -index indexes/lucene-index.neuclir22-zh-en \
  -topics tools/topics-and-qrels/topics.neuclir22-en.original-desc.txt \
  -topicReader TsvInt \
  -output runs/run.neuclir22-zh-en.bm25-default.topics.neuclir22-en.original-desc.txt \
  -bm25 &
bin/run.sh io.anserini.search.SearchCollection \
  -index indexes/lucene-index.neuclir22-zh-en \
  -topics tools/topics-and-qrels/topics.neuclir22-en.original-desc_title.txt \
  -topicReader TsvInt \
  -output runs/run.neuclir22-zh-en.bm25-default.topics.neuclir22-en.original-desc_title.txt \
  -bm25 &

bin/run.sh io.anserini.search.SearchCollection \
  -index indexes/lucene-index.neuclir22-zh-en \
  -topics tools/topics-and-qrels/topics.neuclir22-en.original-title.txt \
  -topicReader TsvInt \
  -output runs/run.neuclir22-zh-en.bm25-default+rm3.topics.neuclir22-en.original-title.txt \
  -bm25 -rm3 &
bin/run.sh io.anserini.search.SearchCollection \
  -index indexes/lucene-index.neuclir22-zh-en \
  -topics tools/topics-and-qrels/topics.neuclir22-en.original-desc.txt \
  -topicReader TsvInt \
  -output runs/run.neuclir22-zh-en.bm25-default+rm3.topics.neuclir22-en.original-desc.txt \
  -bm25 -rm3 &
bin/run.sh io.anserini.search.SearchCollection \
  -index indexes/lucene-index.neuclir22-zh-en \
  -topics tools/topics-and-qrels/topics.neuclir22-en.original-desc_title.txt \
  -topicReader TsvInt \
  -output runs/run.neuclir22-zh-en.bm25-default+rm3.topics.neuclir22-en.original-desc_title.txt \
  -bm25 -rm3 &

bin/run.sh io.anserini.search.SearchCollection \
  -index indexes/lucene-index.neuclir22-zh-en \
  -topics tools/topics-and-qrels/topics.neuclir22-en.original-title.txt \
  -topicReader TsvInt \
  -output runs/run.neuclir22-zh-en.bm25-default+rocchio.topics.neuclir22-en.original-title.txt \
  -bm25 -rocchio &
bin/run.sh io.anserini.search.SearchCollection \
  -index indexes/lucene-index.neuclir22-zh-en \
  -topics tools/topics-and-qrels/topics.neuclir22-en.original-desc.txt \
  -topicReader TsvInt \
  -output runs/run.neuclir22-zh-en.bm25-default+rocchio.topics.neuclir22-en.original-desc.txt \
  -bm25 -rocchio &
bin/run.sh io.anserini.search.SearchCollection \
  -index indexes/lucene-index.neuclir22-zh-en \
  -topics tools/topics-and-qrels/topics.neuclir22-en.original-desc_title.txt \
  -topicReader TsvInt \
  -output runs/run.neuclir22-zh-en.bm25-default+rocchio.topics.neuclir22-en.original-desc_title.txt \
  -bm25 -rocchio &

Evaluation can be performed using trec_eval:

bin/trec_eval -c -m ndcg_cut.20 tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default.topics.neuclir22-en.original-title.txt
python -m pyserini.eval.trec_eval -c -m judged.20 tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default.topics.neuclir22-en.original-title.txt
bin/trec_eval -c -m recall.1000 tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default.topics.neuclir22-en.original-title.txt
bin/trec_eval -c -m map tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default.topics.neuclir22-en.original-title.txt
bin/trec_eval -c -m ndcg_cut.20 tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default.topics.neuclir22-en.original-desc.txt
python -m pyserini.eval.trec_eval -c -m judged.20 tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default.topics.neuclir22-en.original-desc.txt
bin/trec_eval -c -m recall.1000 tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default.topics.neuclir22-en.original-desc.txt
bin/trec_eval -c -m map tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default.topics.neuclir22-en.original-desc.txt
bin/trec_eval -c -m ndcg_cut.20 tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default.topics.neuclir22-en.original-desc_title.txt
python -m pyserini.eval.trec_eval -c -m judged.20 tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default.topics.neuclir22-en.original-desc_title.txt
bin/trec_eval -c -m recall.1000 tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default.topics.neuclir22-en.original-desc_title.txt
bin/trec_eval -c -m map tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default.topics.neuclir22-en.original-desc_title.txt

bin/trec_eval -c -m ndcg_cut.20 tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default+rm3.topics.neuclir22-en.original-title.txt
python -m pyserini.eval.trec_eval -c -m judged.20 tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default+rm3.topics.neuclir22-en.original-title.txt
bin/trec_eval -c -m recall.1000 tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default+rm3.topics.neuclir22-en.original-title.txt
bin/trec_eval -c -m map tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default+rm3.topics.neuclir22-en.original-title.txt
bin/trec_eval -c -m ndcg_cut.20 tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default+rm3.topics.neuclir22-en.original-desc.txt
python -m pyserini.eval.trec_eval -c -m judged.20 tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default+rm3.topics.neuclir22-en.original-desc.txt
bin/trec_eval -c -m recall.1000 tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default+rm3.topics.neuclir22-en.original-desc.txt
bin/trec_eval -c -m map tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default+rm3.topics.neuclir22-en.original-desc.txt
bin/trec_eval -c -m ndcg_cut.20 tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default+rm3.topics.neuclir22-en.original-desc_title.txt
python -m pyserini.eval.trec_eval -c -m judged.20 tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default+rm3.topics.neuclir22-en.original-desc_title.txt
bin/trec_eval -c -m recall.1000 tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default+rm3.topics.neuclir22-en.original-desc_title.txt
bin/trec_eval -c -m map tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default+rm3.topics.neuclir22-en.original-desc_title.txt

bin/trec_eval -c -m ndcg_cut.20 tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default+rocchio.topics.neuclir22-en.original-title.txt
python -m pyserini.eval.trec_eval -c -m judged.20 tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default+rocchio.topics.neuclir22-en.original-title.txt
bin/trec_eval -c -m recall.1000 tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default+rocchio.topics.neuclir22-en.original-title.txt
bin/trec_eval -c -m map tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default+rocchio.topics.neuclir22-en.original-title.txt
bin/trec_eval -c -m ndcg_cut.20 tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default+rocchio.topics.neuclir22-en.original-desc.txt
python -m pyserini.eval.trec_eval -c -m judged.20 tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default+rocchio.topics.neuclir22-en.original-desc.txt
bin/trec_eval -c -m recall.1000 tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default+rocchio.topics.neuclir22-en.original-desc.txt
bin/trec_eval -c -m map tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default+rocchio.topics.neuclir22-en.original-desc.txt
bin/trec_eval -c -m ndcg_cut.20 tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default+rocchio.topics.neuclir22-en.original-desc_title.txt
python -m pyserini.eval.trec_eval -c -m judged.20 tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default+rocchio.topics.neuclir22-en.original-desc_title.txt
bin/trec_eval -c -m recall.1000 tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default+rocchio.topics.neuclir22-en.original-desc_title.txt
bin/trec_eval -c -m map tools/topics-and-qrels/qrels.neuclir22-zh.txt runs/run.neuclir22-zh-en.bm25-default+rocchio.topics.neuclir22-en.original-desc_title.txt

Effectiveness

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

MAP BM25 (default) +RM3 +Rocchio
NeuCLIR 2022 (Chinese): title (original English queries) 0.2658 0.2950 0.2944
NeuCLIR 2022 (Chinese): desc (original English queries) 0.2168 0.2564 0.2489
NeuCLIR 2022 (Chinese): desc+title (original English queries) 0.2750 0.2821 0.2924
nDCG@20 BM25 (default) +RM3 +Rocchio
NeuCLIR 2022 (Chinese): title (original English queries) 0.3705 0.3808 0.3802
NeuCLIR 2022 (Chinese): desc (original English queries) 0.3070 0.3182 0.3206
NeuCLIR 2022 (Chinese): desc+title (original English queries) 0.3723 0.3580 0.3806
J@20 BM25 (default) +RM3 +Rocchio
NeuCLIR 2022 (Chinese): title (original English queries) 0.3908 0.4211 0.4197
NeuCLIR 2022 (Chinese): desc (original English queries) 0.3412 0.3978 0.3974
NeuCLIR 2022 (Chinese): desc+title (original English queries) 0.3899 0.4154 0.4184
Recall@1000 BM25 (default) +RM3 +Rocchio
NeuCLIR 2022 (Chinese): title (original English queries) 0.7567 0.8070 0.8129
NeuCLIR 2022 (Chinese): desc (original English queries) 0.6639 0.7519 0.7404
NeuCLIR 2022 (Chinese): desc+title (original English queries) 0.7567 0.7959 0.8011

Reproduction Log*

To add to this reproduction log, modify this template and run bin/build.sh to rebuild the documentation.