作者: James Allan
DOI: 10.21236/ADA439446
关键词:
摘要: Abstract : The effectiveness of ad-hoc retrieval systems appears to have reached a plateau. After several years 10% gains every year in TREC, improvements dwindled or even stopped. This lack progress was undoubtedly one the reasons behind abandoning suspending TREC after TREC-9. One plausible reason that document has been unable improve is nature task requires adopt "one size fits all" approaches. Given query, system will generally do best return results are good for an "average" user. Doing otherwise (i.e., targeting particular type user) might result substantial on but it just as likely (in environment) cause horrible degradation. By ignoring user (or, more accurately, by treating all users identically), cannot possibly advance beyond level accuracy average specific goal this track bring out hiding, making him her integral part both search process and evaluation. Systems not query chew on, also much information possible about person request, ranging from biographical data, through seeking context, expected result.