
User Profile Identification Based on Text Mining
Various scientific disciplines at the University of West Bohemia in Pilsen attract researchers in various fields of specialty. Because of numerous locations of University buildings combined with overlapping focus of faculties, it is often the case that people of the same interest, working at one institution, do not know each other, which impedes sharing of their experience. It is our objective to merge experts based on similarity of their user profiles. Another objective is to utilize user profiles to find additional documents that may be of interest of a user. Packet filtering approach is used to collect information on users. User profiles are generated with the aid of Suffix Tree Clustering algorithm on the basis of characteristic phrases.
Keywords: text mining, user profile, web, www, recommender system, expert search, clustering, suffix tree, phrase search, characteristic phrase, similarity, packet filter
Year: 2003

Authors of this publication:

Petr Grolmus
E-mail: indy@civ.zcu.cz

Jiřà Hynek
Phone: +420 603492837
E-mail: jhynek@kiv.zcu.cz
WWW: http://www.kiv.zcu.cz/staff/osobni.php?id_osoby=147&lang=EN

Karel Ježek
Phone: +420 377632475
E-mail: jezek_ka@kiv.zcu.cz
WWW: https://cs.wikipedia.org/wiki/Karel_Je%C5%BEek_(informatik)
Related Projects:

Document Classification | |
Authors: | Jiřà Hynek, Karel Ježek, Michal Toman, Roman Tesař, Zdeněk Češka, Petr Grolmus |
Desc.: | Use of inductive machine learning methods in classification of short text documents. |

User Profile Mining, Social Networks | |
Authors: | Jiřà Hynek, Petr Grolmus, Karel Ježek |
Desc.: | Identification of user profiles based on users' behavior on the web. Practical applications in various knowledge and information management projects. |