RSS

Last week I attended the HCIR workshop to present research conducted with Anna Divoli from the University of Chicago earlier this year. This event is a unique opportunity to listen to, socialize and get feedback on your work by the leading people in the field of human-computer information retrieval. 

The Venue
The workshop took place at the Google Mountain View campus in California and included a reception in the evening of the first day and a full day session on the second day. I have been at this campus twice before, but the facilities (and the food!) never cease to impress. Any Google employee from outside of this campus could log into the workshop remotely to watch and listen the presenters, and we could see their engaged faces projected next to the presenters area. I remember from my days as an intern at Google that because of the many interesting events happening nearly daily at some Google campus you needed to choose wisely, which ones to attend.

The Workshop
However, this workshop is not a Google-specific event. Last year it took place at the Rutgers University in New Brunswick, and the year before at SLIS in Washington, DC. Unlike many research conferences, this workshop is organized and attended by people with both academic and industry background. This year there were industry people from Microsoft, Xerox PARC, Elsevier Labs, LinkedIn, as well as smaller companies like Yelp and Glassdoor. With the number of attendees growing each year and reaching approximately 70 people this year, this workshop is certainly a success. Luckily, compared to many other events, the athmosphere is still intimate enough with plently of opportunities to network.   

The Presentations
The workshop started with an introduction, following by a keynote by Prof. Gary Marchionini. I met Gary during my student years, when I volunteered at the JCDL conference at Chapen Hill, North Corolina. At HCIR, Gary talked about the history of search, browsing and information seeking facilitated by computers. Since 1980, the participation of humans has increased dramatically. This is why integrating human intelligence into search is more important than ever. Gary talked also talked about the importance of surrogates, i.e. presentation of search results that makes it easy for people to understand their relevance. It's amazing how incoherent snippets have been dominating search interfaces for decades. He also stressed the need for the availability of high-quality metadata, which drives effective search interfaces. Both of these areas are in immediate focus of Pingar's research. We provide automatically extracted query-based summaries and keywords that replace snippets in our search webpart. Pingar API also allows taxonomy match ups, keyword and entity extraction that produce meaningful content-based metadata. Pingar's metadata add-on for SharePoint that combines this functionality has been popular with customers.    

Pingar at HCIR

Among 60 paper submissions, only 6 were accepted for a full 15-minute presentation and a poster. Pingar's submission was one of these. You can view the slides of the presentation  as well as the poster. You can also view the presentation. It has been invaluable to get feedback on our research with questions from Daniel Tunkelang and Marti Hearst, key people in the field. We are now preparing a full paper to submit to the Special Issue journal describing our work in more detail.   Overall, HCIR has been an excellent workshop and very unique in its organization, focus and coverage among all the ones that I have attended during my academic life. I look forward to preparing another piece of research for next year's workshop.

Comments: