以GAAC分群法提升既有桌面檢索系統之表現－以Google Desktop Search為例

Using the GAAC Clustering Method to Improve the Performance of Desktop Search Engines

魏世杰(Shih-Chieh Wei) ； 周建榮(Chien-Jung Chou)

《Electronic Commerce Studies》 8卷3期 (2010/09) Pp. 323-348

https://doi.org/10.29767/ECS.201009.0002

摘要

傳統桌面檢索系統回傳的文件因為缺乏社群推薦連結之輔助，往往仍需使用者花費心思逐步過濾，才能取得真正所需者。為改善其文件排名，本研究採用特定查詢句分群策略，提出兩階段分群作法。第一階段依底層檢索系統回傳的文件摘要分成兩群，一群排名在前，包含所有查詢句字詞；一群排名在後，包含部分或無查詢句字詞。第二階段就包含所有查詢句的文件，利用群平均聚合分群法(Group-Average Agglomerative Clustering, GAAC)形成群集。再挑出非單一文件群，以其最後結合相似度由高到低決定群間排名。而群內之文件則逐一以最終結合子群之最後結合相似度由高到低決定群內排名。最後，挑出剩餘單一文件群，以其底層檢索系統回傳原始排名順序，遞補於前述分群結果之後，即為調整文件排名。
本研究採用中文標準新聞文件集CIRB030作測試，共49210篇及42個問題集。經由Google Desktop Search回傳原始文件排名結果，再透過雜訊過濾，斷詞、特徵詞選取、建立文件摘要向量。最後以兩階段分群等處理重新排名，再和原始排名做比較。結果顯示經由上述分群調整作法，確實有顯著改善原始桌面檢索系統之排名效果且反應速度快。

關鍵字

GAAC聚合法 ； 特定查詢句分群 ； Google Desktop Search

並列摘要

Traditional desktop search engines such as Google Desktop Search usually return a document ranking which still takes time to filter the desired documents. To improve the document ranking, this work adopts the query-specific clustering approach and proposes a two stage clustering scheme. Based on the returned snippets, the first stage divides the snippets into two groups. The first group contains all keywords in the query and the second group contains partial or no query keywords. The ranking of the first group will be ahead of the second group. In the second stage, the first group is further applied the Group-Average Agglomerative Clustering (GACC) to form hierarchical clusters that all have a combination similarity above a given threshold. Based on the GAAC result, non-singleton clusters are ordered from high to low by their last combination similarity. Within each cluster, the two last combining subclusters are also ordered from high to low by their last combination similarity. Having a combination similarity of 0, singleton clusters will be located behind following their initial snippet order.
As test dataset, a standard Chinese news dataset CIRB030 is used which consists of 49210 documents and 42 enquiry topics. An original document ranking is obtained from Google Desktop Search. Then the snippets are tokenized and filtered to extract the representative keywords and form the snippet vectors. The snippets then go through the two stage clustering scheme to adjust their ranking. The result shows that the two stage clustering scheme can significantly improve the document ranking and the processing time is short.

並列關鍵字

group average agglomerative clustering ； query-specific clustering ； Google Desktop Search

參考文獻

Tsai, C. H. (1996). MMSEG: A word identification system for mandarin Chinese text based on two variants of the maximum matching algorithm. Retrieved from http://technology.chtsai.org/mmseg

中央研究院中文詞知識庫小組(2006)。CIRB030中文新聞語料庫，Retrieved from http://godel.iis.sinica.edu.tw/CKIP/publication.htm。

Chakrabarti, S.(2005).Mining the Web.Morgan Kaufmann Publisher.

Crabtree, D.,Andreae, P.,Gao, X.(2006).Query directed web page clustering.IEEE/WIC/ACM International Conference on Web Intelligence.(IEEE/WIC/ACM International Conference on Web Intelligence).

Cutting, D. R.,Karger, D. R.,Pedersen, J. O.,Tukey, J. W.(1992).Scatter/Gather: A cluster-based approach to browsing large document collections.Proceedings of the 15th ACM SIGIR Conference on Research and Development in Information Retrieval.(Proceedings of the 15th ACM SIGIR Conference on Research and Development in Information Retrieval).