Open Access

Lena Rudenko

• In today's world, data and the information contained therein is one of the most important resources and, at the same time, one of the driving forces behind human development. Not all information and data are equally relevant or important. A huge array of numbers are data too, but without context they are completely meaningless. However, even if these numbers are assigned a certain meaning, it does not automatically make this information useful or interesting for every individual. Thus, personalized information, carefully selected according to the interests and preferences of the individual, is of particular value. The ways in which data and information are transmitted and stored have changed with the development of humanity, from man-to-man legends to social networks where absolutely everyone can create content. Hundreds of millions of Twitter users produce data every second in a continuous stream: they write their own posts, comment, "like" or retweet the posts of the other users,In today's world, data and the information contained therein is one of the most important resources and, at the same time, one of the driving forces behind human development. Not all information and data are equally relevant or important. A huge array of numbers are data too, but without context they are completely meaningless. However, even if these numbers are assigned a certain meaning, it does not automatically make this information useful or interesting for every individual. Thus, personalized information, carefully selected according to the interests and preferences of the individual, is of particular value. The ways in which data and information are transmitted and stored have changed with the development of humanity, from man-to-man legends to social networks where absolutely everyone can create content. Hundreds of millions of Twitter users produce data every second in a continuous stream: they write their own posts, comment, "like" or retweet the posts of the other users, etc. We have an ever-increasing amount of data, which is extremely difficult to navigate. New forms of data require new approaches of their processing and analysis. Stream data processing has received considerable attention in the last decade. The modern applications require the analysis and study of continuous, unstructured and timevarying data instead of static records. One of the most important topics in the field of data processing and analysis is searching for interesting and relevant information for a particular user. The goal of my doctoral thesis is a preference-based analysis of stream data using an efficient algorithm and final aggregation of the resulting tweets. In this work I present a framework that allows the user to reduce the stream of data to only highly relevant information using preferences applied to a wide range of various tweets' attributes. The resulting set of text messages is additionally revised before output, so that the user receives a compact summary and does not have to read a large number of individual tweets.…