Publication: Analysis of Twitter activity by country during competitive events
Loading...
Files
Official URL
Full text at PDC
Publication Date
2018
Advisors (or tutors)
Editors
Journal Title
Journal ISSN
Volume Title
Publisher
Citations
Exportar
Abstract
The subject of this work is an attempt to analyze data. Due to its popularity, availability, and the ability to comment on public events, it was decided to use Twitter data as data for analysis. Data obtained in this way are particularly interesting because of the large social and age cross-section that can be observed among Twitter users, so they give a fairly good statistical sample for a given location. The aim was to create a solution to present the number of tweets posted by users from different countries during media events. Geolocation is usually not included in the tweets, so it was decided to work on the location given by the user [1]. The main tools used in the work are the Twitter API, the non-relational database management system MongoDB and the Python high-level programming language. It was also necessary to use external databases to convert location to coordinates. Thanks to the use of solutions presented in the work, it is possible to present the activity of twitter users from different countries through different visualization methods. In order to analyze the data in the best way, it was decided to use the tweets chart in time, maps showing the total intensity of tweets for the 20 most popular locations for the entire event, and a movie based on maps in minute intervals. It was important for the results to be legible and visually appealing, understandable also for people not related to the project. For the analysis, it was decided to choose popular events on an international scale, the course of which will be possible to reproduce after a while during which it will be possible to observe certain culminating points. During the tests, it was checked how the project deals with two events, the Eurovision Song Contest and the UEFA Champions League match between Real Madrid and Paris Saint Germain, the analysis of the data obtained was also carried out and attempts were made to extract conclusions from them. Thanks to the proposed methods, it was possible to obtain very interesting results, among others, to observe how the culmination such as scoring a goal in a football match influences the activity of users on Twitter and for which nations the given event is the most important.
Description
Universidad Complutense, Facultad de Informática, curso 2017/2018