Distinguished Lecturer Raghu Ramakrishnan: The Future of Search and Information Discovery

Photo of Raghu RamakrishnanRaghu Ramakrishnan, Chief Scientist for Search and Cloud Platforms at Yahoo!, discusses the convergence of browsing and searching on the Internet.  This event on October 19, 2011, is part of the CS department's Distinguished Lecture series.

Ramakrishnan explains that news sites like Yahoo! algorithmically select content to display.  A user's demographic is used to predict which articles are most likely to interest him,  and those are the articles that the site displays first; moreover, the site keeps track of which articles have been clicked on in the past and which have been ignored.  This history is used to perfect the types of articles that are displayed.  Websites implement these techniques in order to maximize a user's CTR (click-through rate), meaning the user is viewing more pages, spending more time on the site, and viewing more advertisements.

Human beings are not altogether removed from the selection of content to be displayed, though.  Editors can review which articles are being selected and which aren't, and they will create future content to match that which is popular now.  Also, these sites can return real-time data about which kinds of people are online, so editors can time the release of content to correspond with times that certain demographics are logging on.  In these ways, sites which were traditionally browsed for content are now doing the browsing for you and showing you what you want to see.

Meanwhile, seaYahoo! logorch engines are evolving so that they no longer simply return a list of URLs whose pages contain the same words as the seach query; rather, the intent of a search is inferred in order to narrow down the field of results.  Ramakrishnan gives the example of searching for "chicken parmigiana Lexington."  Most likely, the person searching for this phrase is looking for a restaurant in Lexington that serves chicken parmigiana, so a modern search engine will only show those types of results, hiding sites that, say, provide chicken parmigiana recipes, or those that sell chicken and parmesan cheese.

Search engines are also beginning to display results in a way that resembles browsing, Ramakrishnan reveals.  A Yahoo! search for "Julia Roberts" returns the normal list of related articles on sites like Wikipedia and IMDb, but first it gives a block of news stories, Internet videos, and recent tweets all related to Julia Roberts.  This format imitates the Yahoo! homepage with its multi-media content and visual appeal, encouraging users to spend more time perusing it.  The pattern of browse results resembling search results and vice versa is, according to Ramakrishnan, likely to continue well into the future.

Bio: Raghu Ramakrishnan is Chief Scientist for Search and Cloud Platforms at Yahoo!, and is a Yahoo! Fellow, heading the Web Information Management research group. His work in database systems, with a focus on data mining, query optimization, and web-scale data management, has influenced query optimization in commercial database systems and the design of window functions in SQL:1999. His paper on the Birch clustering algorithm received the SIGMOD 10-Year Test-of-Time award, and he has written the widely-used text "Database Management Systems" (with Johannes Gehrke). His current research interests are in cloud computing, content optimization, and the development of a "web of concepts" that indexes all information on the web in semantically rich terms. Ramakrishnan has received several awards, including the ACM SIGKDD Innovations Award, the ACM SIGMOD Contributions Award, a Distinguished Alumnus Award from IIT Madras, a Packard Foundation Fellowship in Science and Engineering, and an NSF Presidential Young Investigator Award. He is a Fellow of the ACM and IEEE.