The WWW network is an unlimited source of information, with vast cognitive, scientific or commercial value. Non-structural form, heterogeneity, openness and dynamics of this medium means that there is a need to develop special techniques, algorithms and tools necessary to obtain, filter and output information available on WWW. Brought into being in 2009 Inter-departmental Lab in Web Mining at PJWSTK conducts research and projects in this relatively new field called Web Mining. Lab conducts research that is fundamental as well as concerning latest application, in a cooperation with centres like: Yahoo! Research Barcelona, Gemius S.A., Institute of Fundamental Informatics PAN, Poznań Polytechnics and others.
www: www.users.pjwstk.edu.pl/~msyd/webmining.html
21 June 2010
Inter-departmental Lab in Web Mining
17 June 2010
Visual Microblog
An example of designing of innovative Web 2.0 system that can be used to obtain completely new information on users is a visual microblog. Microblogs (like Twitter) are systems, in which entries are in a form of short information about current state of users. Visual microblog substitutes entries with a choice of one or more combinations from set list of states. Those states are visualised by icons. Consecutive states make up a states history that is represented visually on microblog portal (like a comic book of their own lives). Application that allows to choose a state has been implemented for mobile phones, which allows users for accurate updates of states while writing their own personal microblogs.
Emotions on the Web
16 June 2010
Evaluation of teams of Wikipedia authors
Wikipedia is a very good example of a project that owes its great success to a cooperation of many people without the existence of any formal organisation. Everyone can change the content of its pages. This open model of cooperation without artificial organisational barriers fosters a creative work of teams. Peter Gloor describes in his book Swarm Creativity: Competitive Advantage through Collaborative Innovation Networks the usage of social structure (COIN), in which the coordination is intuitive, without formal leadership as an efficient form of leading the organisation, also in a business sense.
Our research is focused on evaluation of teams editing content of separate articles, based on multidimensional social network. The construction of this network in built on a history of editing all articles and its discussion pages. We have distinguished four dimensions in which we examine relations between authors: trust (based on extending the content created by one author without deleting it), wariness (based on deleting content and restoring the edition), familiarity (modelled on base of activity of pairs of authors on disussion pages), knowledge/interest (linking authors editing content of pages from similar categories).