Organizing Web Search Results Using Clustering By Compression
Abstract
Current Web search engines return long lists of ranked documents that users are forced to sift through to find relevant documents. This paper introduces an interactive presentation method of the search results, based on the notion of clustering by compression. Compression algorithms allow defining a similarity measure based on the degree of common information. Clustering methods allow clustering similar data without any previous knowledge. For this work, we have developed a Java application which retrieves the first 50 results returned by the Google search engine in response to a query, applies some text processing techniques to the Web documents, and thirdly applies the clustering by compression algorithm. The result is a binary tree enhancing the visualization of the formed clusters.