Bio-textual deep web search system based on keyword networks

Abstract

The Internet has been rapidly “deepened” by the prevalence of online databases. With the potentially unlimited information hidden behind their query interfaces, this “Deep Web” of searchable databases is clearly an important frontier for data access. Bioinformatics community has witnessed the growing of deep web and research conducted upon it. Several researches have started upon this relatively unexplored frontier, meta-searcher, database content detection, HTML search form analysis and database raw file merging. Each of them has their own inherited advantages and disadvantages. Based on our new finding of Keyword Networks, we proposed a novel approach to index heterogeneous Deep Web resources. Inheriting form concept of “Six Degrees of Separation”, our implementation has shown its prominent ability of retrieving textual documents from domain specific Deep Web resources. It is capable of integrating state of art deep web contents with minimum human interference and easy to implement on current search engine systems.