Twitter Data Processing

Given the plurality of voices available on the I nternet, it is often difficult to estimate the level of
expertise of the individuals that one encounters in cyber-space. On the other hand, it is also difficult
for any individual expert to establish their credibility in this sea of screaming voices. Therefore, it is
desirable to have an independent mechanism to prove a given person’s claims to authenticity and
legitimacy. One valuable source of information in this endeavour is the individual’s interactions with
content on the web, be it blogging, tweeting, etc. The specific goal of this project is to extract a
collection of topics from the links posted to a given Twitter account to identify the user’s interests and
to use that collection to provide a mechanism for comparing the acuteness of the interest of different
individuals in a given topic.

Faculty Supervisor:

Nando de Freitas

Student:

Partner:

CrowdTrust Technologies Inc

Discipline:

Computer science

Sector:

University:

The University of British Columbia

Program:

Accelerate