For research contributions to information retrieval and human-computer interaction.
ACM-W Names Susan T. Dumais 2014-2015 Athena Lecturer
ACM-W named Susan T. Dumais of Microsoft Research as the 2014-2015 Athena Lecturer. Dumais introduced novel algorithms and interfaces for interactive retrieval that have made it easier for people to find, use and make sense of information. Her research, at the intersection of human-computer interaction and information retrieval, has broad applications for understanding and improving searching and browsing from the Internet to the desktop. The Athena Lecturer award celebrates women researchers who have made fundamental contributions to computer science. It includes a $10,000 honorarium provided by Google Inc.
“Dumais has helped us understand that the search is not the end goal,” said Mary Jane Irwin, who heads the ACM-W awards committee. “Her focus is on understanding when and why people search, and presenting results in context to help integrate those results into the larger search process. Her sustained contributions have shaped the thinking and direction of human-computer interaction and information retrieval, and influenced generations of student interns through collaborative projects with academic and industry partners.”
Dumais’ initial research demonstrated that different people use different vocabulary to describe the same thing, and that this mismatch limits the success of traditional keyword-based information retrieval methods. To build search systems that avoided the vocabulary problem, she and her colleagues invented Latent Semantic Indexing (LSI). A key feature of LSI is its ability to extract the latent conceptual structure from a large collection of texts by analyzing the associations between terms that occur in similar contexts, thus enabling a search engine to retrieve using concepts rather than keywords. Beyond information retrieval, LSI has been used to model various aspects of human cognition such as language acquisition and textual coherence.
Recently, Dumais’ research has analyzed how web content changes over time and how people revisit web pages, establishing that re-visitation patterns are influenced by user intent and changes in content. Her results have produced a retrieval model that uses web page changes to improve search ranking, and new tools to help people understand how the information they interact with changes over time in both expected and unexpected ways. Finally, her research on user modeling and context has enabled search engines to personalize search experiences for different individuals.
The author of more than 200 articles on information science, human-computer interaction, and cognitive science, Dumais holds several patents on novel retrieval algorithms and interfaces.