Web information retrieval inspired by social insect behaviour

This work is focused on retrieving information from Web, which represents the biggest source of information that could be used by human. In the process of retrieving useful information, new approaches have been established by using different principles and theories. Methods of solving problems inspired by nature or biologically are used in Informatics quite frequently. Their potential is confirmed in proposing new and making the well-known methods of solving various kinds of problems more effective. Inspired by social insect’s behaviour and building upon an analysis of related works, we proposed in our work an upgrading of a chosen bee hive model used for retrieving information. The implemented model represents a multiagent system, which uses a bee as a thematic agent for retrieving information from Web. In our work we used an adapted model in order to follow (trace) a story that is developing on the Web that represents a new approach in retrieving information. In this case the time of publishing the story is also important according to the event it is related to. In order to retrieve an increased number of documents of higher quality in the process of searching, we proposed a hierarchical interconnection among several bee hive models. The advantage of the proposed solution in searching with the help of bee hive’s metaphor compared to conventional searching are higher flexibility, more effective source searching that changes during the time and an opportunity to react dynamically to these changes.