Browsed by
Month: May 2017

Going “Deep” into Learning

Going “Deep” into Learning

On Thursday, May 11th, the lab sponsored an exciting tutorial on Deep Learning, or a class of machine learning algorithms for feature extraction and transformation in the realm of artificial intelligence.

The tutorial covered how Brainly, a social Q&A service for students, uses deep learning for new product features including spam detection, question categorization, and finding similar content, to design personalized learning approaches for students. It presented introductory material for people without any previous experience in machine learning, so all in attendance–whether they were beginners or had preexisting knowledge of the topic–took away important information. Overall, we had twenty attendees, and they got a lot more out of the day than a free lunch!

We were honored to have Sashko Zakharchuk as our presenter. After some years in product development and management consulting, Sashko joined Google to work on personalized recommendation systems and text processing algorithms. He is now a machine learning consultant for startups including Brainly in Europe. He delivered a fascinating and informative talk.

The tutorial was a great way to end the semester on a high note for the lab. But don’t worry–we have plenty of tricks up our sleeve for the summer, and we’re continuing to plan for CHIIR 2018 in New Brunswick!

 

We Have a (New) Doctor in the House!

We Have a (New) Doctor in the House!

Long Le, a long time and much loved member of the InfoSeeking Lab, has successfully defended his dissertation, “Extracting Users in Community Question-Answering in Particular Contexts.” Congratulations, Long!

 

Long’s work holds particular import for Community Question-Answering (CQA) sites and their users. He was interested in studying the behavior of the users who participate in CQA. Specifically, he strove to understand how different types of users could be identified based on their behaviors concerning a CQA-specific problem. Rather than discuss users and their actions in a general context, Long extracted contextual situations to develop a more granular analysis of user behavior. Users are the main driving force in CQA and understanding them allows us to know the current state of their respective sites.

 

Obtaining a doctorate is no easy feat, and we’re all incredibly proud of Long and everything he has accomplished. Look out for him in the future–he’ll certainly move forward into big and bright places. Of course, he and his family will also be terribly missed by everyone in the InfoSeeking Lab, but we’re thrilled to count him among our distinguished alumni.