The IBM Haifa Research Lab recently held its Information Retrieval Leadership Seminar, a cutting edge forum that brings together people from academia and industry to share work, exchange ideas, and discuss new directions and trends. Over 100 participants from Israeli universities, startups, and multinational corporations joined together for a full day of lectures, presentations, and demos on the latest technologies for retrieving information.
Large scale information retrieval (IR) is often associated with internet search or with monstrous databases populated with countless emails, legal documents, medical information, images, financial data, and more. Smaller scale IR comes into play for small or medium sized businesses that are also generating massive amounts of information on a daily basis. Various collections of this data must be accessible to people within the company itself-through different forms of enterprise searchwhereas other types of information must be made available to people outside the company.
"With huge volumes of digital information amassing each day, information retrieval and discovery have become more vital than ever before," explained Ronny Lempel, information retrieval expert at the IBM Haifa Labs and organizer of the seminar. "Business and regulatory changes are also contributing to the increased generation of unstructured information, along with the need to search through and analyze this information."
Presentations at the IBM Information Retrieval seminar covered topics such as cluster ranking, text analytics techniques, XML trees and entity-relation graphs, incremental caching, high throughput solutions, and speech information retrieval. The impressive lineup of speakers came from the IBM Haifa Research Lab, Google Inc., the Technion - Israel Institute of Technology, the Max-Planck Institute for Informatics in Germany, ISTI-CNR in Italy, Yahoo! Research in the US, and AOL Relegence Israel.
Two of the seminar's highlights included the keynote address by Andrei Broder from Yahoo! on Algorithmic Advertising and a panel discussion on how Web2.0 is impacting search. Broder discussed the various challenges and calculations required by the search engines so they can customize advertisements for the various searches. The panel, although focused on Web 2.0, revolved around the issues of privacy. Today, search engines know more and more about us and as a result can improve service. However, seminar participants expressed concern that this trend is compromising our privacy and perpetuating a kind of big brother syndrome where our interests and habits have become public knowledge. Some of the questions raised ask whether there is room to expand or restrict the kind of information the search engines can collect and use.
What are the plans for future seminars? The Information Retrieval seminar at the IBM Haifa Research Lab is becoming known as a major event for IR experts in Israel. Each year witnesses a distinct growth in the number of participants and a more elaborate program. "We've been conducting these seminars about once every year and a half and see the audience and diverse program topics as a good combination," noted Lempel. Feedback from the seminar participants was extremely upbeat. Many people attending commented on the high quality content of the presentations and the value they gained learning from and networking with other experts in the field.
"A dedicated conference enables you to jump right into more focused topics and detailed explanations, instead of staying with general superficial discussions," added Lempel. "Bringing together so many people who work on leading edge technologies is a great spark for igniting interesting discussions, joint collaborations, and new ideas."