Integrated text and table mining

Project description

The amount of published scientific articles is growing exponentially, such that researchers are hardly capable of keeping up with advancements in their area. Text mining methods can help to make sense of these large corpora by automatically processing them and extracting relevant information. The need for processing tables in scientific articles is well documented, as they provide a source of succinct and detailed information, such as performance of an approach or details of an experiment. However, this information is often incomplete if considered in isolation, as important aspects necessary to understand the table are elaborated in the full text of the article.

This project aims to develop technologies for hybrid extraction, linking and analysis of tabular and textual data, using a combination of natural language processing, text mining and machine learning techniques. Research would include developing hybrid approaches that jointly operate over unstructured and semi-structured representations (e.g. text and tables), curation of appropriate corpora and their usage in downstream applications such as question answering or semantic retrieval. The project will be placed in the context of different scientific communities, including, for example, (bio)medical, chemical or computer science domains.

The successful candidate will have an excellent first degree in Computer Science or Computational Linguistic, with clear interests and at least some relevant experience in text mining and language modelling. Understanding of machine learning and preferably deep learning is also expected.