InfoPlanet: Visualizing a semantic web to improve search results through exploration and discovery

InfoPlanet visualizes a semantic web that results from users searching on a specific term. This technique helps people explore a dictionary of semantic relationships and discover a set of terms related to their search term. The objectives of the project include the adaptation of suitable ontology to support modeling of general concepts on the web and make those concepts suitable for visual display, the storage and retrieval of those interactively updated concepts based on semantic analytics, and the design of an optimized user interface that enables serendipity and exploration of search terms. It derives ontological and knowledge base information from a lexical database and a commonsense reasoning system, and integrates its own rules to store interconnected concepts. These concepts can be retrieved interactively to a meta-search module to provide interpretation of what the user is looking for and prediction of what the user might be interested in. It also collects user interaction data and gradually learns the semantic relationships among interconnected concepts. Semantic search involves contextual meaning of words and complex motivations of web users, and thus constantly needs verification or clarification from users. Visualization of search query may help to minimize users' effort in such a process. Using InfoPlanet makes searches more effective and enhances the knowledge base of InfoPlanet through successful searches. The project addresses the challenges of designing an intuitive, exploratory tool for web search.