regression.hc4-neuclir22-zh.yaml Maven / Gradle / Ivy

Go to download
---
corpus: neuclir22-zh
corpus_path: collections/multilingual/neuclir22-zh/

index_path: indexes/lucene-index.neuclir22-zh
collection_class: NeuClirCollection
generator_class: DefaultLuceneDocumentGenerator
index_threads: 8
index_options: -storePositions -storeDocvectors -storeRaw -language zh
index_stats:
  documents: 3179206
  documents (non-empty): 3179206

metrics:
  - metric: MAP
    command: tools/eval/trec_eval.9.0.4/trec_eval
    params: -c -m map
    separator: "\t"
    parse_index: 2
    metric_precision: 4
    can_combine: true
  - metric: nDCG@20
    command: tools/eval/trec_eval.9.0.4/trec_eval
    params: -c -m ndcg_cut.20
    separator: "\t"
    parse_index: 2
    metric_precision: 4
    can_combine: false
  - metric: J@20
    command: python -m pyserini.eval.trec_eval
    params: -c -m judged.20
    separator: "\t"
    parse_index: 2
    metric_precision: 4
    can_combine: false
  - metric: Recall@1000
    command: tools/eval/trec_eval.9.0.4/trec_eval
    params: -c -m recall.1000
    separator: "\t"
    parse_index: 2
    metric_precision: 4
    can_combine: false

topic_reader: TsvInt
topic_root: src/main/resources/topics-and-qrels/
qrels_root: src/main/resources/topics-and-qrels/
topics:
  - name: "[HC4 (Chinese): test-topic title](https://github.com/hltcoe/HC4)"
    id: test_title
    path: topics.hc4-v1.0-zh.test.title.tsv
    qrel: qrels.hc4-neuclir22-zh.test.txt
  - name: "[HC4 (Chinese): test-topic description](https://github.com/hltcoe/HC4)"
    id: test_description
    path: topics.hc4-v1.0-zh.test.desc.tsv
    qrel: qrels.hc4-neuclir22-zh.test.txt
  - name: "[HC4 (Chinese): test-topic description+title](https://github.com/hltcoe/HC4)"
    id: test_description_title
    path: topics.hc4-v1.0-zh.test.desc.title.tsv
    qrel: qrels.hc4-neuclir22-zh.test.txt

models:
  - name: bm25-default
    display: BM25 (default)
    params: -bm25 -language zh
    results:
      MAP:
        - 0.0561
        - 0.0428
        - 0.0597
      nDCG@20:
        - 0.0759
        - 0.0687
        - 0.0881
      J@20:
        - 0.0620
        - 0.0590
        - 0.0710
      Recall@1000:
        - 0.4401
        - 0.3565
        - 0.4442
  - name: bm25-default+rm3
    display: +RM3
    params: -bm25 -rm3 -language zh
    results:
      MAP:
        - 0.0453
        - 0.0262
        - 0.0421
      nDCG@20:
        - 0.0630
        - 0.0372
        - 0.0627
      J@20:
        - 0.0540
        - 0.0360
        - 0.0490
      Recall@1000:
        - 0.3996
        - 0.2407
        - 0.2811
  - name: bm25-default+rocchio
    display: +Rocchio
    params: -bm25 -rocchio -language zh
    results:
      MAP:
        - 0.0488
        - 0.0277
        - 0.0462
      nDCG@20:
        - 0.0749
        - 0.0529
        - 0.0735
      J@20:
        - 0.0760
        - 0.0610
        - 0.0740
      Recall@1000:
        - 0.4128
        - 0.3858
        - 0.4259