Matt Lease receives grant from QNRF to improve Arabic language search engine technology

Zhang, Yang  |  Nov 04, 2015

News Image: 
Matt Lease grant from QNRF
Image Caption: 
Matt Lease grant from QNRF
News Image: 
Assistant Professor Matt Lease
Image Caption: 
Matt Lease

While search engines have become incredibly accurate for navigating through websites written in English, finding relevant webpages in other languages is often more difficult.

UT iSchool Associate Professor Matthew Lease and Qatar University Assistant Professor of Computer Science Tamer Elsayed are collaborating together to improve current search engine technology for the Arabic-language Web. Lease and Elsayed received an $884,000 grant from the Qatar National Research Fund for three years for their project “Efficient and Scalable Evaluation for Searching Massive Arabic Social Media and Web Collections.”

“In addition to significantly less research and development investment having been made, the non-English Web is smaller in size for many languages, making it harder to find a relevant needle in a haystack,” Lease said. “Linguistic differences from English can further require tuning search algorithms for each language of interest, and some human populations are inherently polylingual. For example, Arabic is not a single language, but rather a collection of closely-related languages, from Modern Standard Arabic - used for formal writing - to several regional dialects - used in conversation and informal writing.”

To create a controlled environment for search engine experimentation, the professors will crawl the Arabic Web to collect a massive dataset “snapshot”. They will use crowdsourcing to reach Arabic speakers around the world and collect diverse search queries to evaluate the effectiveness of search algorithms developed.

The project also includes significant funding for Lease to fully support doctorate student research assistants as part of his Information Retrieval and Crowdsourcing Research Lab.

“Student research is essential to scientific progress, and I look forward to seeing the amazing things my future research assistants will accomplish on this project,” he said. “It’s been a great pleasure getting to know and help mentor Tamer’s students at Qatar University, and vice versa for him helping mentor students working on the project here at UT-Austin.”

The two professors met at the University of Maryland School of Information while Lease was interviewing for a post-doctorate opportunity and Elsayed was finishing his doctoral degree.

“We were excited to reconnect and renew our iSchool ties across our separate continents,” Lease said. “This project idea provided the perfect opportunity to work together on a problem of mutual interest which is of great practical importance to society and presents us with plenty of tough technical challenges to make it all work.”

 

glqxz9283 sfy39587stf02 mnesdcuix8
sfy39587stf03
sfy39587stp14