Project contact: jaak.vilo (at) stacc.ee

Social network analysis

Social and communication networks have enjoyed a tremendous surge in popularity recently, drawing together friends, acquaintances, colleagues and other people with similar interests or backgrounds.These networks also produce a major portion of Internet content services, such as blogs, photo and video archives, and interest group forums.Web marketing and user-oriented advertising are gaining considerable ground too.The aim of social network analysis is to explore methods for analysing global networks, their structure, nature and evolution (generative models); to identify specialised sub-networks; and to predict changes in user behaviour for better or worse.This knowledge can be used to offer new services to users faster and more conveniently, while preventing criminal and malicious activities like spam and scams.The main client of the project is Skype, which has a network of half a billion users.

Web log data mining

Web log data mining draws on the fact that all web, mobile and computer-based software programmes save log entries about users.The challenge is how to identify typical users and their usage habits from logs and to compare the actual use of software applications to the initial visions of the software developers.The project aims to develop a program that can save and import usage logs into a data warehouse and use cluster analysis to identify user groups, in order to visualise and describe different usage patterns so as to help software designers in making decisions.For instance, user interfaces could be optimised and modified to allow users to use them the way they want.

Predicting user behaviour

While web log data mining works with logs of past activities, the latest trends, and typical user groups and interests, then predicting user behaviour anticipates the behaviour of users in soft real-time.Its goal is to explore solutions for automatic user analysis by identifying the very latest trends and user interests and by adapting user interfaces accordingly, thus providing a better user experience.

STACC News