Methods for Thematic Analysis and Ranking in Information Analysis Systems Based on Relational DBMS

The paper presents the models and algorithms for searching, filtering and ranking data in an information analysis system, based on the ontological description of its structure. The rules for filtering and ranking objects based on their boolean, numeric or enumerable properties are specified by the users, and based on these rules SQL queries are evaluated. Another problem considered is taking subject area into account when searching for objects. Adding this allows one to solve many tasks that occur in current research information systems, such as searching for experts in given subject areas or generating reports. A search query can be specified in the form of a set of keywords, and a statistical approach is used for comparing keyword sets.