期刊名称：International Journal of Computer Trends and Technology
电子版ISSN：2231-2803
出版年度：2013
卷号：4
期号：9-4
出版社：Seventh Sense Research Group
摘要：Web search applications represent user information needs by submission of query to search engine. But still the entire query submitted to search engine doesn’t satisfy the user information needs, because users may want to get information on diverse aspects when they submit the same query. From this discovering the numeral of dissimilar user search goals for query and depicting each goal with several keywords automatically become complicated. The suggestion and examination of user search goals can be very valuable in improving search engine importance and user knowledge. Discovering the numeral of dissimilar user search goals for query by kmeans clustering with user feedback sessions. Efficiently reflect user information needs generate a pseudodocument to map the different user feedback sessions. Clustering Pseudo documents with K means clustering result are computationally difficult and semantic similarity between the pseudo terms is also important while clustering. To conquer this problem proposed a FCM clustering algorithm to group the pseudo documents and it also measure the semantic similarity between the pseudo terms in the documents using wordnet. The FCM algorithm divides pseudo documents data for dissimilar size cluster by using fuzzy systems. FCM choosing cluster size and central point depend on fuzzy model. The FCM clustering algorithm it congregate quickly to a local optimum or grouping of the pseudo documents in wellorganized way. Semantic similarity between the pseudo terms with Wordnet based similarity is used for comparing the similarity and diversity of pseudo terms. Finally experimental result measures the clustering results with parameters like classified average precision (CAP), Voted AP (VAP), risk to avoid classifying search results and average precision (AP). It shows FCM based system improve the feedback sessions outcome than the normal pseudo documents.
关键词：User search goals; feedback sessions; pseudo-documents; classified average precision (CAP); Voted AP (VAP); average precision (AP); Fuzzy C means clustering; K-means clustering
