Research:Predict Users Search

Created

2014/11

Contact

Guillaume Sabran: guilaumesabran@fas.harvard.edu. Affiliation: Harvard University

Collaborators

Kevin Kuate Fodouop: kuatefodouop@fas.harvard.edu

Duration:  2014-11 – 2014-12

 Wikimedia supported

We would need data related to pages users visited (in English), when they visited those pages and if possible if it was through direct link within Wikipedia. Also, data about the users (gender, age, country...) might relevant be we are not sure at this point.

---

Research:Projects

This page documents a planned research project.
Information may be incomplete and change before the project starts.

Key Personnel

• John Doe
• jane Doe

Methods

We would analyse the relations between the Wikipedia pages graph and users' search and then try to make predictions that could be used as suggestions if relevant

Dissemination

This is a class based project. I don't think there is any requirement to keep results/code private and if so we would share what is of interest. If results are of particular importance, we might publish a paper.

Benefits for the Wikimedia community

If results are good, this would let the Wikipedia fondation have a better understanding of the way people look for knowledge. It could also enhance users experience if relevant recommendations were made to them.

Timeline

We would like to start working on this as shortly as possible. This is a project for a Machine Learning class me and Kevin are taking at MIT. The project would end in mid December.