Formal concept analysis based web pages classification/visualization and their application to information retrieval

The net users are required a lot of time to treat with web search engines due to their retrieved results as ranking lists. In order to reduce load of users, `search@once' that provides well-structured outputs as line-diagram, is proposed. The proposed system employs formal concept analysis to classify web pages and visualize the lattice structure of web pages and keywords as line diagram. This system is implemented on the computer (CPU=2.83GHz! $MM=2GB), by using Python, which is an object-oriented programming language, Application Program Interface (API), and one of the GUI libraries, Tkinter. Through the subjective evaluation and sign test with seven subjects, it is confirmed that search@once is helpful to understand the overview of the retrieved results and to find new keywords for refining the results.