Seminars

View all Seminars  |  Download ICal for this event

Unified Question Answering over RDF Knowledge Graphs and Natural Language Text

Series: Department Seminar

Speaker: Dr. Soumajit Pramanik, IIT Bhilai.

Date/Time: Dec 19 11:00:00

Location: CSA Seminar Hall (Room No. 254, First Floor)

Faculty Advisor:

Abstract:
Question answering over knowledge graphs and other RDF data has been greatly advanced, with a number of good systems providing crisp answers for natural language questions or telegraphic queries. Some of these systems incorporate textual sources as additional evidence for the answering process, but cannot compute answers that are present in text alone. Conversely, systems from the IR and NLP communities have addressed QA over text, but such systems barely utilize semantic data and knowledge. In this work, we develop a QA system that can seamlessly operate over RDF datasets and text corpora, or both together, in a unified framework. Our method, called Uniqorn, builds a context graph on-the-fly, by retrieving question-relevant triples from the RDF data and/or snippets from a text corpus, using a fine-tuned BERT model. The resulting graph is typically rich but highly noisy. Uniqorn copes with this input by advanced graph algorithms for Group Steiner Trees, that identify the best answer candidates in the context graph. Experimental results on several benchmarks of complex questions with multiple entities and relations, show that Uniqorn produces results comparable to the state-of-the-art on KGs, text corpora, and heterogeneous sources. The graph-based methodology provides user-interpretable evidence for the complete answering process.

Speaker Bio:
Dr. Soumajit Pramanik is an Assistant Professor in the Department of Electrical Engineering and Computer Science (EECS) at Indian Institute of Technology Bhilai. Prior to joining IIT Bhilai, he was a Postdoctoral researcher at the Max Planck Institute for Informatics, Saarbrücken, Germany. Dr. Pramanik's research interest lies in the fields of Information Retrieval, Machine Learning, Social Computing and Complex Networks. His doctoral work was solely devoted to the progress of our understanding of multilayer networks and their applications. Dr. Pramanik identified four particularly important problems related to structural and functional properties of multilayer networks - (a) information diffusion, (b) community detection, (c) node movement and (d) entity recommendation and investigated each of them thoroughly. During his Postdoctoral research, he worked on complex question-answering over hybrid knowledge bases and textual corpora. He (along with his collaborators) has developed QUEST, a method that can answer complex questions directly from textual sources on-the-fly, by computing similarity joins over partial results from different documents. He is currently working on extending this work for answering factoid questions jointly over both knowledge graphs and text corpora. Moreover, he has also started exploring the areas of conversational and temporal question answering.

Host Faculty: R Govindarajan