Ontology based Topic Model for analyzing text in Nanomaterial Informatics

Level: 30hp (Advanced level, Master's thesis)
Number of students: 1-2
Area: Computer Science

Topic models have been widely and successfully used for a number of text mining tasks such as information retrieval and sentiment analysis. They support a way to find semantics in the text. However, there are still difficulties for humans to understand topics as they depend on word distributions from the text. Some researchers have realized that an ontology containing the semantics in a domain could help people to interpret the results of topic models. Further, from the perspective of an ontology engineer, latent semantics obtained from topic models could also bring benefits for ontologies for example in ontology enrichment and ontology learning.

This thesis work is part of a larger research project on materials design and aims to interpret the result of a topic model for nanomaterials by using a nanoparticle ontology as a background knowledge or to enrich and learn a nanoparticle ontology by using the result of a topic model for nanomaterials. The project will be based on an existing framework of ontology-based topic models over nanomaterial data. The framework will be implemented and the use for interpretation of topic models and enrichment of an ontology will be evaluated.

Prerequisites:
It is an advantage to have knowledge about ontologies and/or topic models.

Contact: Huanyu Li, Patrick Lambrix.

Page responsible: Patrick Lambrix
Last updated: 2018-10-29