Dr. Brian E. Ulicny, senior scientist, and Dr. Mieczyslaw M. Kokar, president, Versatile Information Systems Inc., Framingham, Mass., will receive approximately $450,000 in funding for the 3-year project entitled "Automated Ontologically-Based Link Analysis of International Web Logs for the Timely Discovery of Relevant and Credible Information."

"It can be challenging for information analysts to tell what's important in blogs unless you analyze patterns," Ulicny said. Patterns include the content of the blogs as well as what hyperlinks are contained within the blog.

Within blogs, hyperlinks act like reference citations in research papers thereby allowing someone to discover the most important events bloggers are writing about in just the same way that one can discover the most important papers in a field by finding which ones are the most cited in research papers.

This type of analysis can help information analysts' searches be as productive as possible.

The blog study is part of Air Force Office of Scientific Research's new Information Forensics and Process Integration research program recently launched at Syracuse University, Syracuse, N.Y.

The new portfolio of projects consists of three areas of research emphasis - incomplete information and metrics; search, interactive design, and active querying; and cognitive processing.

One of the problems analysts may have with blog monitoring, Ulicny noted, is there is too much actionable information for the analyst to properly analyze.

"We are developing an automated tool to tell analysts what bloggers are most interested in at a point in time," Ulicny said. This analysis, Kokar said, is based on what Versatile Information Systems calls the RSTC approach to blog analysis - relevance, specificity, timeliness, and credibility. RSTC helps information analysts filter the most important information to study.

"Relevance involves developing a point of focus and information related to a particular focus," Kokar said. Timeliness has to do with immediacy - how important is a topic now. "Credibility," he continued, "is the amount of trust you have in an information source."

Finally, specificity can provide value to information analysts depending on how general or specific they need the information to be.

In some ways, the team's automated project works like a search engine but with a more focused approach. Traditional search engines present users with information based on, for example, the number of times a term appears in a document.

The information obtained via a search engine query tends to be similar among the documents returned. Blog postings, however, can be much more dissimilar from one to another.

"What we're doing is a sort of information retrieval," Ulicny said. "The difference is that in order to find and analyze blog entries, you need to more adequately model how the blogs work on a global scale."

To some degree blog interpretation, he said, involves understanding a different form of communication.

"Blog entries have a different structure," Ulicny said. "They are typically short and are about something external to the blog posting itself , such as a news event. It's not uncommon for a blogger to simply state, 'I can't believe this happened,' and then link to a news story."

In this example, Ulicny said, there might not be much of interest in the blog posting, yet the fact that the blogger called attention to this story can be significant to understanding what matters.

A good example, he said, is the recent furor in the Muslim world over the publication of cartoons of Mohammad in a Danish newspaper. The original publication wasn't much noticed in the West, but bloggers discussed this event that possibly contributed to riots worldwide.

"The fact that the web is a vast source of information is sometimes overlooked by military analysts," Kokar said. "Our research goal is to provide the warfighter with a kind of information radar to better understand the information battlespace."

Related Links
Air Force Office of Scientific Research