Open GoogleCodeExporter opened 9 years ago
Had the chance to look at this briefly in the context of the plural/singular
problem:
HIVE is relying on Lucene for indexing and search. Lucene is indexing both the
preferred and alternate labels. "Noxious mammals" includes the alternate labels
"Harmful mammals" and "Injurious mammals", which appear to be increasing its
score.
Test case:
1. Select vocabulary: Agrovoc
2. Search term: "mammals"
3. Results:
Noxious mammals (9.35)
Mammals (6.64)
Game mammals (4.15)
Aquatic mammals (4.15)
Expected results:
Mammals should appear first. We need to consider what to expect for the remaining results (perhaps factor in relationships -- is narrower/broader than exact match; whether phrase is in altLabel, etc).
Original comment by craig.wi...@unc.edu
on 18 Apr 2011 at 5:24
Original issue reported on code.google.com by
rscherle
on 17 Feb 2011 at 6:57