ITU Paper To Be Published In Prestigious EMNLP 2015


ITU Paper To Be Published In Prestigious EMNLP 2015
July 30, 2015

The Data Science lab of Information Technology University is pleased to share yet another commendable milestone. One of their research papers entitled “An Unsupervised Method for Discovering Lexical Variations in Roman Urdu Informal Text” has been recently accepted for publication at Empirical Methods on Natural Language Processing (EMNLP 2015), Lisboa Portugal.  With the advent of burgeoning growth of Social media and SMS, the trend of Roman Urdu usage has increased radically. The prime motivation behind this paper was to standardize Roman Urdu and form a framework for its analysis to derive meaningful insights. EMNLP is highly prestigious venue (CORE ‘A’ flag bearer and H-Index =  56 over Google Metrics) in the area of natural language processing and was very competitive this year with the acceptance rate of mere 24% (312 acceptance out of 1315 submissions).

Another startling aspect about this achievement is that this is the work of ITU MS student Abdul Qayyum and Data Science Lab RA Moeen Uddin. They worked under the supervision of Dr. Faisal Kamiran, Director Data Science Lab of ITU who has more than 20 research publications including 8 A*/A ones with 257 citations and 8 H-index. They have prototyped the outcome of their research efforts in the form of an android app which is available for download at PlayStore   (https://play.google.com/store/apps/details?id=com.balytix.romanurdudictionary). ”