PLAN 80: Article
New Faculty: Sepandar Kamvar

World Leader in Social Computing and Information Management

Sepandar Kamvar, considered one of the best young minds in the growing field of social computing, has been appointed associate professor in media arts and sciences beginning January 1.

Previously a consulting assistant professor at Stanford’s Institute for Computational and Mathematical Engineering, Kamvar was the founder and CEO of Kaltix, a personalized search engine acquired by Google in 2003, then led the personalization effort at Google from 2003-2007.

‘Sep’s work combines design with analysis, theory with practice and depth with breadth,’ says Mitchel Resnick, head of the Media Arts and Sciences academic program. ‘We see him as a perfect match for the culture of the Media Lab.’

For his part, Kamvar says: ‘MIT, and the Media Lab in particular, offers a rich intellectual environment unconstrained by traditional disciplinary boundaries. I'm happy to be joining a great set of colleagues and students at an exciting time in the Lab.’

In the field of information retrieval – historically most concerned with queries and documents – Kamvar has developed models that focus on the people behind the queries and documents. More specifically, he’s devised search technologies where questions are answered by people rather than documents (social search), that extract and summarize information on people (social information mining), where results are dependent not just on the query but also on the person making the query (personalized search) and where searches are modeled on human social norms for reputation and sharing (peer-to-peer search). His research contributions can be divided into four main categories:

Personalized search. As a graduate student, Kamvar developed the first efficient algorithm for adding personal context to the internet search process, enabling customized searches to match the needs of each individual; the algorithm is now part of Google’s basic search software, used by more than a billion people each week.

Reputation algorithms. He devised a reputation mechanism that computes the level of ‘trust’ for individuals in peer-to-peer networks so that more reliable individuals can be given higher rankings in search results; his research paper on the subject has been cited more than 2500 times, making it one of the most highly-cited papers in social computing.

Social search. He helped develop algorithms for locating people in a network who are best able to provide the requested information, rather than locating documents that provide the information; he also developed strategies to encourage participation in such networks, which have had an impact on a number of technologies including the startup company Aardvark.

Emotion mining. He has developed search engines to analyze and visualize human emotion using the words and pictures that people post on their social networks as raw material for creating mosaics of humanity; this work has served as the basis for art installations at several major museums including the Museum of Modern Art in New York, the Victoria and Albert Museum in London and the National Museum of Contemporary Art in Athens.

Kamvar is the author of two books and over 40 technical publications and patents in the fields of search and social computing, and serves on the advisory boards of several companies, including Clever Sense and Etsy. He holds a PhD from Stanford in scientific computing and computational mathematics (2004) and a BS in chemistry from Princeton (1999).