Choosing Conference Papers

Client

Kilian Weinberger, Associate Professor, Department of Computer Science
Email: <kilianweinberger@cornell.edu>

Student contact

James Russo <jdr289@cornell.edu> is setting up a team for this project. If you are interested in joining the team, please contact him.

Background

A big problem at academic conferences is for attendees to find out which talks or poster presentations to look at. This project attempts to solve this problem.

Most academics have Google scholar profiles, e.g. https://scholar.google.com/citations?user=jsxk8vsAAAAJ&hl=en, which list all the publications of a researcher. The project will allow academic researchers to provide a link to their Google scholar profile and their email address, and to select which conference they are interested in. All papers in the Google scholar profile are downloaded and compared to the publications at the conference. The user will receive an email with a list of the most similar papers published at the conference and those that cite his or her past work.

Project Summary

From a technical point of view the project requires the design and implementation of all parts of the system. This is likely to include: an interactive front end accessed from a web page that communicates and populates a database with publications; integration with a machine learned similarity metric between academic publications; and email to send results to the user. The machine learning metric has already been developed by a separate research project.