Skip to content

Commit

Permalink
Mrtydi regression (#1702)
Browse files Browse the repository at this point in the history
Add regression and test case for all other Mr. TyDi languages, following #1685. 

including: 
- template and yaml files (index stats, MRR@100 and Recall@100)
- test cases (size of topics and qrels, doc loading)
- unrelated to Mr.TyDi: add the path to `/bin/build.sh` in the readme
  • Loading branch information
crystina-z committed Dec 15, 2021
1 parent f782151 commit 1514044
Show file tree
Hide file tree
Showing 116 changed files with 57,535 additions and 7 deletions.
3 changes: 2 additions & 1 deletion README.md
Original file line number Diff line number Diff line change
Expand Up @@ -79,7 +79,7 @@ For the most part, these runs are based on [_default_ parameter settings](https:
+ Regressions for [CLEF 2006 Monolingual French](docs/regressions-clef06-fr.md)
+ Regressions for [TREC 2002 Monolingual Arabic](docs/regressions-trec02-ar.md)
+ Regressions for FIRE 2012: [Monolingual Bengali](docs/regressions-fire12-bn.md), [Monolingual Hindi](docs/regressions-fire12-hi.md), [Monolingual English](docs/regressions-fire12-en.md)
+ Regressions for Mr. TyDi: [ar](docs/regressions-mrtydi-v1.1-ar.md)
+ Regressions for Mr. TyDi: [ar](docs/regressions-mrtydi-v1.1-ar.md), [bn](docs/regressions-mrtydi-v1.1-bn.md), [en](docs/regressions-mrtydi-v1.1-en.md), [fi](docs/regressions-mrtydi-v1.1-fi.md), [id](docs/regressions-mrtydi-v1.1-id.md), [ja](docs/regressions-mrtydi-v1.1-ja.md), [ko](docs/regressions-mrtydi-v1.1-ko.md), [ru](docs/regressions-mrtydi-v1.1-ru.md), [sw](docs/regressions-mrtydi-v1.1-sw.md), [te](docs/regressions-mrtydi-v1.1-te.md), [th](docs/regressions-mrtydi-v1.1-th.md)

## Reproduction Guides

Expand Down Expand Up @@ -131,6 +131,7 @@ If you've found Anserini to be helpful, we have a simple request for you to cont
In the course of [reproducing](docs/reproducibility.md) baseline results on standard test collections, please let us know if you're successful by sending us a pull request with a simple note, like what appears at the bottom of [the Robust04 page](docs/regressions-robust04.md).
Reproducibility is important to us, and we'd like to know about successes as well as failures.
Since the regression documentation is auto-generated, pull requests should be sent against the [raw templates](https://github.com/castorini/anserini/tree/master/src/main/resources/docgen/templates).
Then the regression documentation can be generated using the [`bin/build.sh`](bin/build.sh) script.
In turn, you'll be recognized as a [contributor](https://github.com/castorini/anserini/graphs/contributors).

Beyond that, there are always [open issues](https://github.com/castorini/anserini/issues) we would appreciate help on!
Expand Down
66 changes: 66 additions & 0 deletions docs/regressions-mrtydi-v1.1-bn.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,66 @@
# Anserini: Regressions for [Mr. TyDi (Bengali)](https://github.com/castorini/mr.tydi)

This page documents regression experiments for [Mr. TyDi (Bengali)](https://github.com/castorini/mr.tydi).

The exact configurations for these regressions are stored in [this YAML file](../src/main/resources/regression/mrtydi-v1.1-bn.yaml).
Note that this page is automatically generated from [this template](../src/main/resources/docgen/templates/mrtydi-v1.1-bn.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 MrTyDiCollection \
-input /path/to/mrtydi-v1.1-bn \
-index indexes/lucene-index.mrtydi-v1.1-bengali.pos+docvectors+raw \
-generator DefaultLuceneDocumentGenerator \
-threads 1 -storePositions -storeDocvectors -storeRaw -language bn \
>& logs/log.mrtydi-v1.1-bn &
```

See [this page](https://github.com/castorini/mr.tydi) for more details about the Mr. TyDi corpus.
For additional details, see explanation of [common indexing options](common-indexing-options.md).

## Retrieval

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

```
nohup target/appassembler/bin/SearchCollection -index indexes/lucene-index.mrtydi-v1.1-bengali.pos+docvectors+raw \
-topicreader TsvInt -topics src/main/resources/topics-and-qrels/topics.mrtydi-v1.1-bn.train.txt.gz \
-output runs/run.mrtydi-v1.1-bn.bm25.topics.mrtydi-v1.1-bn.train.txt.gz \
-language bn -bm25 -hits 100 &
nohup target/appassembler/bin/SearchCollection -index indexes/lucene-index.mrtydi-v1.1-bengali.pos+docvectors+raw \
-topicreader TsvInt -topics src/main/resources/topics-and-qrels/topics.mrtydi-v1.1-bn.dev.txt.gz \
-output runs/run.mrtydi-v1.1-bn.bm25.topics.mrtydi-v1.1-bn.dev.txt.gz \
-language bn -bm25 -hits 100 &
nohup target/appassembler/bin/SearchCollection -index indexes/lucene-index.mrtydi-v1.1-bengali.pos+docvectors+raw \
-topicreader TsvInt -topics src/main/resources/topics-and-qrels/topics.mrtydi-v1.1-bn.test.txt.gz \
-output runs/run.mrtydi-v1.1-bn.bm25.topics.mrtydi-v1.1-bn.test.txt.gz \
-language bn -bm25 -hits 100 &
```

Evaluation can be performed using `trec_eval`:

```
tools/eval/trec_eval.9.0.4/trec_eval -c -M 100 -m recip_rank -c -m recall.100 src/main/resources/topics-and-qrels/qrels.mrtydi-v1.1-bn.train.txt runs/run.mrtydi-v1.1-bn.bm25.topics.mrtydi-v1.1-bn.train.txt.gz
tools/eval/trec_eval.9.0.4/trec_eval -c -M 100 -m recip_rank -c -m recall.100 src/main/resources/topics-and-qrels/qrels.mrtydi-v1.1-bn.dev.txt runs/run.mrtydi-v1.1-bn.bm25.topics.mrtydi-v1.1-bn.dev.txt.gz
tools/eval/trec_eval.9.0.4/trec_eval -c -M 100 -m recip_rank -c -m recall.100 src/main/resources/topics-and-qrels/qrels.mrtydi-v1.1-bn.test.txt runs/run.mrtydi-v1.1-bn.bm25.topics.mrtydi-v1.1-bn.test.txt.gz
```

## Effectiveness

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

MRR@100 | BM25 |
:---------------------------------------|-----------|
[Mr. TyDi (Bengali): train](https://github.com/castorini/mr.tydi)| 0.3566 |
[Mr. TyDi (Bengali): dev](https://github.com/castorini/mr.tydi)| 0.3385 |
[Mr. TyDi (Bengali): test](https://github.com/castorini/mr.tydi)| 0.4182 |


R@100 | BM25 |
:---------------------------------------|-----------|
[Mr. TyDi (Bengali): train](https://github.com/castorini/mr.tydi)| 0.8336 |
[Mr. TyDi (Bengali): dev](https://github.com/castorini/mr.tydi)| 0.8432 |
[Mr. TyDi (Bengali): test](https://github.com/castorini/mr.tydi)| 0.8694 |
66 changes: 66 additions & 0 deletions docs/regressions-mrtydi-v1.1-en.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,66 @@
# Anserini: Regressions for [Mr. TyDi (English)](https://github.com/castorini/mr.tydi)

This page documents regression experiments for [Mr. TyDi (English)](https://github.com/castorini/mr.tydi).

The exact configurations for these regressions are stored in [this YAML file](../src/main/resources/regression/mrtydi-v1.1-en.yaml).
Note that this page is automatically generated from [this template](../src/main/resources/docgen/templates/mrtydi-v1.1-en.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 MrTyDiCollection \
-input /path/to/mrtydi-v1.1-en \
-index indexes/lucene-index.mrtydi-v1.1-english.pos+docvectors+raw \
-generator DefaultLuceneDocumentGenerator \
-threads 1 -storePositions -storeDocvectors -storeRaw -language en \
>& logs/log.mrtydi-v1.1-en &
```

See [this page](https://github.com/castorini/mr.tydi) for more details about the Mr. TyDi corpus.
For additional details, see explanation of [common indexing options](common-indexing-options.md).

## Retrieval

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

```
nohup target/appassembler/bin/SearchCollection -index indexes/lucene-index.mrtydi-v1.1-english.pos+docvectors+raw \
-topicreader TsvInt -topics src/main/resources/topics-and-qrels/topics.mrtydi-v1.1-en.train.txt.gz \
-output runs/run.mrtydi-v1.1-en.bm25.topics.mrtydi-v1.1-en.train.txt.gz \
-language en -bm25 -hits 100 &
nohup target/appassembler/bin/SearchCollection -index indexes/lucene-index.mrtydi-v1.1-english.pos+docvectors+raw \
-topicreader TsvInt -topics src/main/resources/topics-and-qrels/topics.mrtydi-v1.1-en.dev.txt.gz \
-output runs/run.mrtydi-v1.1-en.bm25.topics.mrtydi-v1.1-en.dev.txt.gz \
-language en -bm25 -hits 100 &
nohup target/appassembler/bin/SearchCollection -index indexes/lucene-index.mrtydi-v1.1-english.pos+docvectors+raw \
-topicreader TsvInt -topics src/main/resources/topics-and-qrels/topics.mrtydi-v1.1-en.test.txt.gz \
-output runs/run.mrtydi-v1.1-en.bm25.topics.mrtydi-v1.1-en.test.txt.gz \
-language en -bm25 -hits 100 &
```

Evaluation can be performed using `trec_eval`:

```
tools/eval/trec_eval.9.0.4/trec_eval -c -M 100 -m recip_rank -c -m recall.100 src/main/resources/topics-and-qrels/qrels.mrtydi-v1.1-en.train.txt runs/run.mrtydi-v1.1-en.bm25.topics.mrtydi-v1.1-en.train.txt.gz
tools/eval/trec_eval.9.0.4/trec_eval -c -M 100 -m recip_rank -c -m recall.100 src/main/resources/topics-and-qrels/qrels.mrtydi-v1.1-en.dev.txt runs/run.mrtydi-v1.1-en.bm25.topics.mrtydi-v1.1-en.dev.txt.gz
tools/eval/trec_eval.9.0.4/trec_eval -c -M 100 -m recip_rank -c -m recall.100 src/main/resources/topics-and-qrels/qrels.mrtydi-v1.1-en.test.txt runs/run.mrtydi-v1.1-en.bm25.topics.mrtydi-v1.1-en.test.txt.gz
```

## Effectiveness

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

MRR@100 | BM25 |
:---------------------------------------|-----------|
[Mr. TyDi (English): train](https://github.com/castorini/mr.tydi)| 0.1592 |
[Mr. TyDi (English): dev](https://github.com/castorini/mr.tydi)| 0.1685 |
[Mr. TyDi (English): test](https://github.com/castorini/mr.tydi)| 0.1404 |


R@100 | BM25 |
:---------------------------------------|-----------|
[Mr. TyDi (English): train](https://github.com/castorini/mr.tydi)| 0.5785 |
[Mr. TyDi (English): dev](https://github.com/castorini/mr.tydi)| 0.6196 |
[Mr. TyDi (English): test](https://github.com/castorini/mr.tydi)| 0.5365 |
66 changes: 66 additions & 0 deletions docs/regressions-mrtydi-v1.1-fi.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,66 @@
# Anserini: Regressions for [Mr. TyDi (Finnish)](https://github.com/castorini/mr.tydi)

This page documents regression experiments for [Mr. TyDi (Finnish)](https://github.com/castorini/mr.tydi).

The exact configurations for these regressions are stored in [this YAML file](../src/main/resources/regression/mrtydi-v1.1-fi.yaml).
Note that this page is automatically generated from [this template](../src/main/resources/docgen/templates/mrtydi-v1.1-fi.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 MrTyDiCollection \
-input /path/to/mrtydi-v1.1-fi \
-index indexes/lucene-index.mrtydi-v1.1-finnish.pos+docvectors+raw \
-generator DefaultLuceneDocumentGenerator \
-threads 1 -storePositions -storeDocvectors -storeRaw -language fi \
>& logs/log.mrtydi-v1.1-fi &
```

See [this page](https://github.com/castorini/mr.tydi) for more details about the Mr. TyDi corpus.
For additional details, see explanation of [common indexing options](common-indexing-options.md).

## Retrieval

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

```
nohup target/appassembler/bin/SearchCollection -index indexes/lucene-index.mrtydi-v1.1-finnish.pos+docvectors+raw \
-topicreader TsvInt -topics src/main/resources/topics-and-qrels/topics.mrtydi-v1.1-fi.train.txt.gz \
-output runs/run.mrtydi-v1.1-fi.bm25.topics.mrtydi-v1.1-fi.train.txt.gz \
-language fi -bm25 -hits 100 &
nohup target/appassembler/bin/SearchCollection -index indexes/lucene-index.mrtydi-v1.1-finnish.pos+docvectors+raw \
-topicreader TsvInt -topics src/main/resources/topics-and-qrels/topics.mrtydi-v1.1-fi.dev.txt.gz \
-output runs/run.mrtydi-v1.1-fi.bm25.topics.mrtydi-v1.1-fi.dev.txt.gz \
-language fi -bm25 -hits 100 &
nohup target/appassembler/bin/SearchCollection -index indexes/lucene-index.mrtydi-v1.1-finnish.pos+docvectors+raw \
-topicreader TsvInt -topics src/main/resources/topics-and-qrels/topics.mrtydi-v1.1-fi.test.txt.gz \
-output runs/run.mrtydi-v1.1-fi.bm25.topics.mrtydi-v1.1-fi.test.txt.gz \
-language fi -bm25 -hits 100 &
```

Evaluation can be performed using `trec_eval`:

```
tools/eval/trec_eval.9.0.4/trec_eval -c -M 100 -m recip_rank -c -m recall.100 src/main/resources/topics-and-qrels/qrels.mrtydi-v1.1-fi.train.txt runs/run.mrtydi-v1.1-fi.bm25.topics.mrtydi-v1.1-fi.train.txt.gz
tools/eval/trec_eval.9.0.4/trec_eval -c -M 100 -m recip_rank -c -m recall.100 src/main/resources/topics-and-qrels/qrels.mrtydi-v1.1-fi.dev.txt runs/run.mrtydi-v1.1-fi.bm25.topics.mrtydi-v1.1-fi.dev.txt.gz
tools/eval/trec_eval.9.0.4/trec_eval -c -M 100 -m recip_rank -c -m recall.100 src/main/resources/topics-and-qrels/qrels.mrtydi-v1.1-fi.test.txt runs/run.mrtydi-v1.1-fi.bm25.topics.mrtydi-v1.1-fi.test.txt.gz
```

## Effectiveness

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

MRR@100 | BM25 |
:---------------------------------------|-----------|
[Mr. TyDi (Finnish): train](https://github.com/castorini/mr.tydi)| 0.4101 |
[Mr. TyDi (Finnish): dev](https://github.com/castorini/mr.tydi)| 0.4133 |
[Mr. TyDi (Finnish): test](https://github.com/castorini/mr.tydi)| 0.2836 |


R@100 | BM25 |
:---------------------------------------|-----------|
[Mr. TyDi (Finnish): train](https://github.com/castorini/mr.tydi)| 0.8198 |
[Mr. TyDi (Finnish): dev](https://github.com/castorini/mr.tydi)| 0.8285 |
[Mr. TyDi (Finnish): test](https://github.com/castorini/mr.tydi)| 0.7193 |
66 changes: 66 additions & 0 deletions docs/regressions-mrtydi-v1.1-id.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,66 @@
# Anserini: Regressions for [Mr. TyDi (Indonesian)](https://github.com/castorini/mr.tydi)

This page documents regression experiments for [Mr. TyDi (Indonesian)](https://github.com/castorini/mr.tydi).

The exact configurations for these regressions are stored in [this YAML file](../src/main/resources/regression/mrtydi-v1.1-id.yaml).
Note that this page is automatically generated from [this template](../src/main/resources/docgen/templates/mrtydi-v1.1-id.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 MrTyDiCollection \
-input /path/to/mrtydi-v1.1-id \
-index indexes/lucene-index.mrtydi-v1.1-indonesian.pos+docvectors+raw \
-generator DefaultLuceneDocumentGenerator \
-threads 1 -storePositions -storeDocvectors -storeRaw -language id \
>& logs/log.mrtydi-v1.1-id &
```

See [this page](https://github.com/castorini/mr.tydi) for more details about the Mr. TyDi corpus.
For additional details, see explanation of [common indexing options](common-indexing-options.md).

## Retrieval

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

```
nohup target/appassembler/bin/SearchCollection -index indexes/lucene-index.mrtydi-v1.1-indonesian.pos+docvectors+raw \
-topicreader TsvInt -topics src/main/resources/topics-and-qrels/topics.mrtydi-v1.1-id.train.txt.gz \
-output runs/run.mrtydi-v1.1-id.bm25.topics.mrtydi-v1.1-id.train.txt.gz \
-language id -bm25 -hits 100 &
nohup target/appassembler/bin/SearchCollection -index indexes/lucene-index.mrtydi-v1.1-indonesian.pos+docvectors+raw \
-topicreader TsvInt -topics src/main/resources/topics-and-qrels/topics.mrtydi-v1.1-id.dev.txt.gz \
-output runs/run.mrtydi-v1.1-id.bm25.topics.mrtydi-v1.1-id.dev.txt.gz \
-language id -bm25 -hits 100 &
nohup target/appassembler/bin/SearchCollection -index indexes/lucene-index.mrtydi-v1.1-indonesian.pos+docvectors+raw \
-topicreader TsvInt -topics src/main/resources/topics-and-qrels/topics.mrtydi-v1.1-id.test.txt.gz \
-output runs/run.mrtydi-v1.1-id.bm25.topics.mrtydi-v1.1-id.test.txt.gz \
-language id -bm25 -hits 100 &
```

Evaluation can be performed using `trec_eval`:

```
tools/eval/trec_eval.9.0.4/trec_eval -c -M 100 -m recip_rank -c -m recall.100 src/main/resources/topics-and-qrels/qrels.mrtydi-v1.1-id.train.txt runs/run.mrtydi-v1.1-id.bm25.topics.mrtydi-v1.1-id.train.txt.gz
tools/eval/trec_eval.9.0.4/trec_eval -c -M 100 -m recip_rank -c -m recall.100 src/main/resources/topics-and-qrels/qrels.mrtydi-v1.1-id.dev.txt runs/run.mrtydi-v1.1-id.bm25.topics.mrtydi-v1.1-id.dev.txt.gz
tools/eval/trec_eval.9.0.4/trec_eval -c -M 100 -m recip_rank -c -m recall.100 src/main/resources/topics-and-qrels/qrels.mrtydi-v1.1-id.test.txt runs/run.mrtydi-v1.1-id.bm25.topics.mrtydi-v1.1-id.test.txt.gz
```

## Effectiveness

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

MRR@100 | BM25 |
:---------------------------------------|-----------|
[Mr. TyDi (Indonesian): train](https://github.com/castorini/mr.tydi)| 0.2972 |
[Mr. TyDi (Indonesian): dev](https://github.com/castorini/mr.tydi)| 0.2937 |
[Mr. TyDi (Indonesian): test](https://github.com/castorini/mr.tydi)| 0.3762 |


R@100 | BM25 |
:---------------------------------------|-----------|
[Mr. TyDi (Indonesian): train](https://github.com/castorini/mr.tydi)| 0.7948 |
[Mr. TyDi (Indonesian): dev](https://github.com/castorini/mr.tydi)| 0.7827 |
[Mr. TyDi (Indonesian): test](https://github.com/castorini/mr.tydi)| 0.8426 |
Loading

0 comments on commit 1514044

Please sign in to comment.