Here are a couple of items that might make for good additions to your
search and IR reading lists.
+ From the Stanford Database Group
Leveraging Context to Resolve Identity in Photo Albums
Since tagging images is a popular topic these days, I thought this new paper was worthy of your attention. You’ve heard of PageRank but now you can read about PeopleRank.
From the abstract:
Our system suggests likely identity labels for photographs in a personal photo collection. Instead of using face recognition techniques, the system leverages automatically-available context, like the time and location where the photos were taken. Based on time and location, the system automatically computes event and location groupings of photos. As the user annotates some of the identities of people in their collection, patterns of re-occurrence and co-occurrence of different people in different locations and events emerge. The system uses these patterns to generate label suggestions for identities that were not yet annotated. These suggestions can greatly accelerate the process of manual annotation.
+ Web searching made more successful with automated, personalized assistance system
This news release presents highlights of a paper by Dr. Jim Jansen at Penn State University. The paper is currently available online (abstract free, full text fee-based) and will be available in print this summer.
A Penn State researcher has developed software that improves Web searching with a personalized system that offers automated assistance for structuring and refining queries, evaluating search results and finding more relevant information.
“Research shows 50 percent of all Web results retrieved are not relevant, pointing to a need for improved searching techniques,” said Jim Jansen, assistant professor of information sciences and technology. “This technology enabled a 20-percent performance increase.” Other approaches to personalizing searches rely upon “explicit feedback” where the system interrupts searchers as they hunt for information. Research has shown that only 1 percent to 2 percent of users are likely to use such systems because of the extra effort involved, Jansen said.
We blogged about other research by Dr. Jansen last October.