What we do
This project is devoted to a research on social algorithms and studying effects which they have on reality (in particular, within reputation, recommendatory and collaborative systems). Similar algorithms, on the one hand, are focused on reflection of real estimates of reputation, a contribution (for example, for reputation and collaborative systems), the importance or preferences (for the formation of recommendations). On another hand, widespread introduction of such algorithms leads to changes in reality and formation of the behavior aimed at receiving an appreciation, or implicit restriction of availability of sources of information because of features of their ranging. As a result, the constructive research of similar algorithms, including empirical researches, has significant scientific prospects, as from the estimation of the influence of algorithms on reality as from the development of these algorithms around social mechanisms and practices.
Thus, within this project we plan to explore a number of hypotheses:
- accounting of behavioral factors of users will allow to increase efficiency of recommendations and improve received estimates
- the structural factors revealed in the analysis of communities by methods of machine learning and the analysis of social networks influence efficiency of interaction and mutual reputation assessment of participants (users of the system)
- identification of mismatch of rating estimates and feedbacks will allow to reveal shortcomings of the existing rating systems (TripAdvisor cases, comments at Stack Exchange) and offer ways of their improvement
- different classes of "points of interest" (PoI) in ratings have different level of reputation "attenuation"
Have you spotted a typo?
Highlight it, click Ctrl+Enter and send us a message. Thank you for your help!
To be used only for spelling or punctuation mistakes.