Search

Document collections resulting from searches in the biomedical literature, for instance, in PubMed, are often so large that some organization of the returned information is necessary. Clustering is an efficient tool for organizing search results. To help the user to decide how to continue the search for relevant documents, the content of each cluster can be characterized by a set of representative keywords or cluster labels. As different users may have different interests, it can be desirable with solutions that make it possible to produce labels from a selection of different topical categories. We therefore introduce the concept of multi-focus cluster labeling to give users the possibility to get an overview of the contents through labels from multiple viewpoints. The concept for multi-focus cluster labeling has been established and has been demonstrated on three different document collections. We illustrate that multi-focus visualizations can give an overview of clusters along axes that general labels are not able to convey. The approach is generic and should be applicable to any biomedical (or other) domain with any selection of foci where appropriate focus vocabularies can be established. A user evaluation also indicates that such a multi-focus concept is useful. [ABSTRACT FROM AUTHOR]

Clustering of search result is undoubtedly a tool that can provide the summarization of the millions of documents in a way where a user can easily locate his/her information. To guide user to the right cluster of documents, cluster labels should be meaningful and correctly representing the clusters. However significant a cluster is, if the label is not proper, user will never select it. In this paper, we present a method to label clusters based on their linking information. Our cluster labeling method is independent of any clustering method but restricted to only search result documents. We use heuristic search method to find all the linked documents of a cluster. If all or some documents of a cluster share hyperlinks, then we deduce label from these linked documents’ titles using famous Apriori algorithm for frequent itemset mining. This removes the requirement of reviewing other members of a cluster in labeling process. [ABSTRACT FROM AUTHOR]