|
Identifier |
000388642 |
Title |
Using Twitter to study correlation between nutrition and health |
Author
|
Καλαϊτζάκης, Ανδρέας Γεώργιος Μιχαήλ Κ.
|
Thesis advisor
|
Noha, Ibrahim
|
Abstract |
The last two decades we became witnesses of a rapid development of distributed computing and
computer networks. Users that were initially restricted to access static text data that was available on
the Internet are now enjoying multimedia content that is even produced by other users in real-time. The
increasing proliferation and affordability of Internet devices, as well as the ease of publishing,
searching and accessing information on the web encourages the individual users to communicate their
content with the web society giving birth to the idea of social interaction imposing a growing need for
systems that can extract useful information from this amount of data. One of the fundamental problems
that emerged in social media stream analysis with a wide range of applications is to effectively detect
underlying topics and their associated documents. It becomes clear that modern social services and
social media show a substantial potential of providing society with a rather promising source of
information which prevails over the traditional ones on a series of important dimensions. Recruiting
social media in order to inform the public has proved to have a significantly lower operating cost in
conjunction with a better propagation velocity. These advantages encouraged the academic community
to investigate a framework under which a partial replacement of the traditional sentinel surveillance
services with web enhanced ones could take place. Mobilized by this emerging need and recognizing a
significant void in empirical studies that focus on nutrition, we randomly collected more than 200
millions tweets along with a series of accompanying features in a two months' period. Applying state of
the art text analysis techniques on the aforementioned dataset we were able to draw significant
conclusions on the dynamics that characterize the sentinel related traffic focusing mainly on well-being
aspects that are related with nutrition patterns within the population.
|
Language |
English |
Subject |
Causal relations |
|
Topic analysis |
Issue date |
2014-11-21 |
Collection
|
School/Department--School of Sciences and Engineering--Department of Computer Science--Post-graduate theses
|
|
Type of Work--Post-graduate theses
|
Views |
343 |