Knowledge extraction and organization to support question-answering

The project aims at developing a question answering system for the National Bank of Canada, which can be used to answer the questions by users in the domain of banking and financial services, as well as human resources. The NBC currently uses a generic QA system from which some limitations have been observed, namely the limited coverage of domain-specific entities and concepts and low accuracy of the question answering. The project will develop and test approaches to solve these problems, by collecting more concepts in the domain and organizing them into a knowledge graph. The question answering system will then leverage the domain-specific knowledge graph.

Faculty Supervisor:

Jian-Yun Nie

Student:

Partner:

Banque Nationale du Canada

Discipline:

Computer science

Sector:

Finance and Insurance; Professional, scientific and technical services

University:

Université de Montréal

Program: