Skip to content

Latest commit

 

History

History
92 lines (62 loc) · 6.17 KB

regressions-robust04.md

File metadata and controls

92 lines (62 loc) · 6.17 KB
Raw

Anserini: Regressions for Disks 4 & 5 (Robust04)

This page describes regressions for the TREC 2004 Robust Track, which uses TREC Disks 4 & 5. 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.

Indexing

Typical indexing command:

nohup sh target/appassembler/bin/IndexCollection -collection TrecCollection -input /path/to/robust04 \
 -index lucene-index.robust04.pos+docvectors+rawdocs -generator DefaultLuceneDocumentGenerator -threads 16 \
 -storePositions -storeDocvectors -storeRaw >& log.robust04.pos+docvectors+rawdocs &

The directory /path/to/disk45/ should be the root directory of TREC Disks 4 & 5; inside each there should be subdirectories like ft, fr94. Note that Anserini ignores the cr folder when indexing, which is the standard configuration.

For additional details, see explanation of common indexing options.

Retrieval

Topics and qrels are stored in src/main/resources/topics-and-qrels/, downloaded from NIST:

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

nohup target/appassembler/bin/SearchCollection -index lucene-index.robust04.pos+docvectors+rawdocs \
 -topicreader Trec -topics src/main/resources/topics-and-qrels/topics.robust04.txt \
 -bm25 -output run.robust04.bm25.topics.robust04.txt &

nohup target/appassembler/bin/SearchCollection -index lucene-index.robust04.pos+docvectors+rawdocs \
 -topicreader Trec -topics src/main/resources/topics-and-qrels/topics.robust04.txt \
 -bm25 -rm3 -output run.robust04.bm25+rm3.topics.robust04.txt &

nohup target/appassembler/bin/SearchCollection -index lucene-index.robust04.pos+docvectors+rawdocs \
 -topicreader Trec -topics src/main/resources/topics-and-qrels/topics.robust04.txt \
 -bm25 -axiom -axiom.deterministic -rerankCutoff 20 -output run.robust04.bm25+ax.topics.robust04.txt &

nohup target/appassembler/bin/SearchCollection -index lucene-index.robust04.pos+docvectors+rawdocs \
 -topicreader Trec -topics src/main/resources/topics-and-qrels/topics.robust04.txt \
 -qld -output run.robust04.ql.topics.robust04.txt &

nohup target/appassembler/bin/SearchCollection -index lucene-index.robust04.pos+docvectors+rawdocs \
 -topicreader Trec -topics src/main/resources/topics-and-qrels/topics.robust04.txt \
 -qld -rm3 -output run.robust04.ql+rm3.topics.robust04.txt &

nohup target/appassembler/bin/SearchCollection -index lucene-index.robust04.pos+docvectors+rawdocs \
 -topicreader Trec -topics src/main/resources/topics-and-qrels/topics.robust04.txt \
 -qld -axiom -axiom.deterministic -rerankCutoff 20 -output run.robust04.ql+ax.topics.robust04.txt &

Evaluation can be performed using trec_eval:

eval/trec_eval.9.0.4/trec_eval -m map -m P.30 src/main/resources/topics-and-qrels/qrels.robust04.txt run.robust04.bm25.topics.robust04.txt

eval/trec_eval.9.0.4/trec_eval -m map -m P.30 src/main/resources/topics-and-qrels/qrels.robust04.txt run.robust04.bm25+rm3.topics.robust04.txt

eval/trec_eval.9.0.4/trec_eval -m map -m P.30 src/main/resources/topics-and-qrels/qrels.robust04.txt run.robust04.bm25+ax.topics.robust04.txt

eval/trec_eval.9.0.4/trec_eval -m map -m P.30 src/main/resources/topics-and-qrels/qrels.robust04.txt run.robust04.ql.topics.robust04.txt

eval/trec_eval.9.0.4/trec_eval -m map -m P.30 src/main/resources/topics-and-qrels/qrels.robust04.txt run.robust04.ql+rm3.topics.robust04.txt

eval/trec_eval.9.0.4/trec_eval -m map -m P.30 src/main/resources/topics-and-qrels/qrels.robust04.txt run.robust04.ql+ax.topics.robust04.txt

Effectiveness

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

MAP BM25 +RM3 +Ax QL +RM3 +Ax
TREC 2004 Robust Track Topics 0.2531 0.2903 0.2896 0.2467 0.2747 0.2774
P30 BM25 +RM3 +Ax QL +RM3 +Ax
TREC 2004 Robust Track Topics 0.3102 0.3365 0.3333 0.3079 0.3232 0.3229

Replication Log