Research Activities

Here are current / past research activities in brief :
  • Intelligent Tutoring System: I was a member of DeepTutor project. The project developed an advanced intelligent tutoring system that fostered students' deep understanding of complex science topics through quality interaction and instruction. It used Learning Progressions (LPs), a recently developed framework by the science education research community to describe students' natural paths to mastery. I have published several papers (see the publications tab) on semantic textual similarity, pronoun resolution, question and dialog generation in the context of intelligent tutoring system.
  • Personal Assistant Dialog Systems :  Due to the advancement in automatic speech recognition technologies, personal assistant dialog systems such as Apple's Siri, Google Now, Microsoft Cortana and Nuance Dragon Go are now increasingly a part of our daily life. Such systems need to identify the intentions of users' queries and provide answers to their intentions. I've done very effective internships at Microsoft Research and AT&T Labs Research in this direction that resulted publications and patents.
  • Review Mining:  The advent of web 2.0 technology has brought about an exponential growth of internet content. Users actively participate on information exchange by writing their views on the items they bought or movies they watch and so on. Analysing such reviews is of great value not only to the other users but also to the merchant. We have published a paper that presents a system to extract and aggregate analogous features in Reviews. 
  • Reference ReconciliationTo design a semantic information integration system, we are faced to two reconciliation problems. First, the schema (or ontology) reconciliation which consists in finding mappings between elements (concepts or relations) of two schemas or two ontologies. The second problem concerns data reconciliation (named reference reconciliation) which consists in comparing data descriptions and deciding whether different descriptions refer to the same real world entity (e.g. the same person, the samearticle, the same gene).  I've worked on the reference reconciliation problem while I was working in LEO Team of INRIA, France. We developed a tool, called LN2R,  to reconcile references. We participated on OAEI 2010 conference. The results are very inspiring. Details can be found on the paper.
  • Ontology Matching : As mentioned earlier, ontology matching aka schema matching is needed as a first step in information integration. To address this issue, we have developed a tool, called TaxoMap,  in LEO Team of INRIA. We have participated in OAEI 2009 and OAEI 2010 conference with this tool. Details are provided in respective papers. 
  • Semantic P2P data management: I've developed P2P infrastructure required by  "SomeWhere", a semantic peer-to-peer data management system. 
  • Video Streaming in Mobile Ad Hoc Network : Video streaming is very challenging in Mobile Ad Hoc network. In this research work, we have proposed CLAPS , a P2P solution to video streamingin MANET. Details can be found in this paper.