Automatic mapping of concepts and relationships

Summary

This project aims to investigate ways to generate concept maps automatically from documents.

Supervisor(s)

Dr Rafael A. Calvo

Research Location

Electrical and Information Engineering

Program Type

Synopsis

Concept maps are graphical tools used to represent knowledge; they map concepts and their relationships in a hierarchical way, making it easier to understand complex topics. They are seldom used due to the effort required to generate them. Generating concept maps automatically from documents would allow a visual approach to summarizing documents automatically. We are developing machine algorithms to model the documents semantic properties. These algorithms use vector representation of the documents (called ‘bag-of-words’ models) and feature selection algorithms that reduce the problem’s dimensionality.

Keywords

text-mining, computer supported collaborative learning, computer supported cooperative work, concept-maps, knowledge maps

Opportunity ID

The opportunity ID for this research opportunity is: 276

Other opportunities with Dr Rafael A. Calvo

• Semantic awareness in collaborative writing
• Social network visualization in collaborative work and learning
• Evaluation of classification algorithms for Brain Computer Interfaces
• Multimodal emotion recognition
• Multimodal approaches for detecting attention in Human-Computer Interaction