Sample Blocking Rules for Falcon Data Sets

Sample blocking rule sequences

Products

  • Blocking rule sequence 1

    • modelno_title_jaccard < 0.14 AND title_dice < 0.33 AND price_reldiff >= 0.74 AND modelno_title_overlap < 0.5

    • modelno_title_cosine < 0.23 AND modelno_overlap < 0.5 AND brand_jaccard < 0.13 AND modelno_title_dice < 0.06 AND title_dice < 0.61

  • Blocking rule sequence 2

    • title_overlap < 1.5 AND modelno_title_cosine < 0.41 AND title_jaccard < 0.22 AND price_reldiff >= 0.28

    • title_dice < 0.17 AND modelno_jaccard < 0.8 AND title_jaccard < 0.71 AND pcategory1_dice < 0.14

Songs

  • Blocking rule sequence 1

    • artist_name_jaccard < 0.29

    • title_jaccard < 0.94 AND release_dice < 0.94

    • title_jaccard < 0.94 AND title_dice < 0.97 AND title_dice3g < 0.8

  • Blocking rule sequence 2

    • release_jaccard < 0.41

    • title_dice < 0.96 AND title_jaccard < 0.94

Citations

  • Blocking rule sequence 1

    • journal_jaccard < 0.29

    • title_jaccard < 0.94 AND authors_dice < 0.94

    • title_jaccard < 0.94 AND title_dice < 0.97 AND title_dice3g < 0.8

  • Blocking rule sequence 2

    • title_dice < 0.61 AND title_jaccard < 0.57

    • title_jaccard < 0.32

    • authors_dice < 0.49 AND authors_jaccard < 0.14

    • authors_dice < 0.13 AND title_overlap < 5.5