What are you searching for?
Information retrieval seminar introduces the hottest new trends in search technology and the biggest challenges coming our way.
A recent seminar organized by IBM Research – Haifa gathered close to 150 people of Israel's information retrieval community. Why all the interest? According to David Konopnicki, this year's seminar organizer and manager of information retrieval solutions at the Haifa Lab, search and information retrieval are vital to extract value from the masses of information being produced and collected today by all kinds of industries.
Information retrieval, or IR, analyzes how we access information and the automatic processes used by computers to find relevant information. Some describe it as the science of searching for documents or information, whether the search is conducted among your personal documents or for more general information on the web.
According to Konopnicki, IR is at the forefront of both research and industry because 'search' is now big business.
"With so much textual information available, good search technologies and text analytics can help us understand and retrieve what's out there," he explained. "And it doesn't matter whether it's genetic researchers searching for correlations between patients and treatments, market analysts looking for trends, financial mavens seeking data, or students doing their homework on geography projects. Advances in information retrieval benefit humanity as a whole."
IBM's new Watson computer, scheduled to compete against top Jeopardy! contestants in February, is just one example of how advanced search technologies are being used to help computers figure 'think' and 'understand' what information is relevant to the question at hand. Although distinguishing between relevant and irrelevant information is something humans do in an instant, getting computers to identify valuable data is a complex challenge.
"Today's information retrieval community is facing new challenges, including very large-scale text analytics, alongside the application of search technologies to new areas such as customer-behavior analytics or advertisements," explained Konopnicki.
The Israeli IR community apparently agrees. The local talent showed up in full force to enjoy lectures from experts from four different universities in Israel – including the Technion, University of Haifa, Ben Gurion University, and Tel Aviv University – alongside presentation from industry leaders such as NICE and IBM.
Mining Twitter and semantic advertising
This year's seminar focused on text analytics technologies, with exciting new applications like Twitter mining and relevance to advertisement. Keynote speaker Prof. Oren Etzioni, Director of the Turing Center at the University of Washington, fascinated the crowd with his talk "Open Information Extraction at Web Scale."
The event offered participants a chance to discuss common challenges and innovative solutions. The seminar featured a good balance of talks on theoretical work and research, actual applications in the industry, and fascinating demos from various companies.
What's the next big thing is information retrieval? Konopnicki sees it as search without query - search based only on your context and the computer knowing what you are looking for without you asking the question.
"Soon, you'll find yourself walking down the street and automatically get information on your mobile device suggesting you stop by a nearby bookstore that's having a special," Konopnicki said. "That won't be random information; it'll all be based on your current location, the free time on your calendar, and books your might have ordered in the past."
In short, soon our computers will find the answers before we even ask the questions. It's all a matter of what you're searching for.