Document collections resulting from searches in e.g. the PubMed literature 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 key words or cluster labels. Solutions based on adapted state-of-the-art methodology, have through the project been developed and integrated into the CoreMine system, to enable clustering of results from keyword searched in PubMed and providing textual labels and summaries describing the contents of the clusters.  

However, studies of manual label assignment show that the choice of labels is subjective and will be dependent on each person’s judgment, preferences and interests. As a solution to this we have therefore introduced the concept of multi-focus cluster labeling giving users the possibility to get an overview of the contents through labels from multiple viewpoints. This can also provide views into the document collection along other axes than the clustering does, giving multiple views into the same set without reclustering.