Canada Spends Logo

Explore our Budget 2026 analysis

Grant: Discovery Grants Program - Individual

École Polytechnique de Montréal

$46,000.00

A Semantic Parsing Approach to Open-world Knowledge Graph Question Answering

FY 2023-24

Summary
Knowledge bases, an important artifact of Artificial intelligence and Semantic Web, have become ubiquitous. In the form of knowledge graphs, they also represent an important backbone of information retrieval on the Web. However, they are not easily accessible to laypeople, as they require a specific understanding of query languages such as SPARQL and of knowledge base schemas.  Being able to perform knowledge base question answering (KBQA) with natural language has been a long-standing goal of artificial intelligence, as it involves natural language understanding, knowledge representation and reasoning and the ability to map natural language with knowledge base elements. The long-term objective of this research program is to design and develop multi-modal semantic search engines that exploit knowledge bases / graphs and pretrained language models through a natural language interface. The short-term objectives, in this proposal, are to explore a generative approach to knowledge base question answering and reasoning through various neural architectures for (SPARQL) query generation, question representation, and neural reasoning. We particularly aim at designing generalizable models, which integrate schema embeddings and are able to adapt to various knowledge bases and various questions (complex questions, questions or queries unseen in training, multilingual questions) in incomplete settings. Such types of settings require knowledge base completion methods and automatic reasoning. This research program will train 8 HQP in advanced natural language processing, knowledge base and representation learning methods and neural models. It will have an impact at the fundamental level (natural language understanding, knowledge-based artificial intelligence) and at the practical level, with the development of natural language-powered interfaces for question answering on the Semantic Web. Such interfaces can benefit various sectors (culture, education, health, .) and the general public.
Program
Discovery Grants Program - Individual
Location
Quebec, Canada
Recipient Type
Academia
Type
Grant
Source
NSERC's Awards Data

As published

Application id
RGPIN-2023-05634
Name nom
Zouaq, Amal
Department département
Génie informatique et génie logiciel - Génie informatique et génie logiciel
Organization id
64
Institution établissement
École Polytechnique de Montréal
Province en
Québec
Province fr
Québec
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2023
Competition year année de concours
2023
Award amount
46000
Program id
RGPIN
Program name en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
DISCOVERY RESEARCH
Group fr
RECHERCHE AXÉE SUR LA DÉCOUVERTE
Committee code
1507
Committee name en
Computer Science
Committee name fr
Sciences informatiques
Area of application code
1207
Area of application group en
Advancement of knowledge
Area of application group fr
Avancement des connaissances
Area of application en
Information, computer and communication technologies
Area of application fr
Technologie de l'information, des ordinateurs et des réseaux de communications
Research subject code
2803
Research subject group en
Artificial intelligence
Research subject group fr
Knowledge representation
Research subject en
Intelligence artificielle
Research subject fr
Répresentation des connaissances
Application title
A Semantic Parsing Approach to Open-world Knowledge Graph Question Answering
Keyword
Knowledge Base Linking; Multi-hop question answering; Neural Reasoning; Query Generation; Schema Integration
Application summary
Knowledge bases, an important artifact of Artificial intelligence and Semantic Web, have become ubiquitous. In the form of knowledge graphs, they also represent an important backbone of information retrieval on the Web. However, they are not easily accessible to laypeople, as they require a specific understanding of query languages such as SPARQL and of knowledge base schemas.  Being able to perform knowledge base question answering (KBQA) with natural language has been a long-standing goal of artificial intelligence, as it involves natural language understanding, knowledge representation and reasoning and the ability to map natural language with knowledge base elements. The long-term objective of this research program is to design and develop multi-modal semantic search engines that exploit knowledge bases / graphs and pretrained language models through a natural language interface. The short-term objectives, in this proposal, are to explore a generative approach to knowledge base question answering and reasoning through various neural architectures for (SPARQL) query generation, question representation, and neural reasoning. We particularly aim at designing generalizable models, which integrate schema embeddings and are able to adapt to various knowledge bases and various questions (complex questions, questions or queries unseen in training, multilingual questions) in incomplete settings. Such types of settings require knowledge base completion methods and automatic reasoning. This research program will train 8 HQP in advanced natural language processing, knowledge base and representation learning methods and neural models. It will have an impact at the fundamental level (natural language understanding, knowledge-based artificial intelligence) and at the practical level, with the development of natural language-powered interfaces for question answering on the Semantic Web. Such interfaces can benefit various sectors (culture, education, health, .) and the general public.
École Polytechnique de Montréal - $46,000 Grant (2023… | Canada Spends