|
I have a new experimental version of OTS which I developed in Java. The new algorithm implements NLP (natural language processing) techniques which were not used in the C++ version of OTS. This version uses the same XML dictionaries as the regular version.
This demo shows the feature of grouping articles of similar content together. The program does not use the article's source or title to sort it by. Each group of articles is given a title that describes the content of the group. In the demo, the user loads articles in HTML and TXT format from the hard-drive and results from a Google web search. Next the user tells the program to group the articles together. The program scans the articles and sorts them by their topics. Each Folder is then given a title by the program. The title describes the topics covered by the different articles. |