A search of the collection of documents generally provide excerpts of the documents are arranged according to rank matches in a long list. Not infrequently a search result in tens and even hundreds of fragments of documents that caused a user to scroll the screen up and down (scrolling) to examine the documents snippet one by one. This situation causes a user is having difficulty in determining which documents relevant to the topic he wants. In this Final Project developed an application web-based document segmentation with suffix tree clustering method. The basic concept of this method is to classify documents in the search results to form groups or clusters based on words or phrases contained in these documents. The application requires the search input and output will result in clusters containing the corresponding documents. This cluster can be stratified depending on the word or phrase that might be distinguished on the same parent cluster. Clusters generated is displayed to the user. Then on the last cluster is selected will display a collection of documents, each consisting of the title and snippet of the document. With this method expected results would be easier to trace. Keywords : text mining, suffix tree, suffix tree clustering, the grouping of documents.
展开▼
