– Make use of complex and extensive information.
To quickly grasp and form an idea about a research field, a subject area, or a comprehensive investigation material is a challenge for many professions. What is worth investigating further, what does science say, what seems to be connected?
At this Lunch & Learn, we showcase an example of combining visualization and semantic analysis using machine learning to tackle the above challenge.
Gustaf Nelhans and Johan Eklund from the Data as Impact Lab at Högskolan i Borås talk about how they first used bibliometric methods to map and analyze relevant literature and investigate the content and metadata. They then proceeded to do semantic analyses on the identified articles. They developed a strategy for extracting phrases that could constitute answers to specific questions. The approach, known as question answering (QA), uses deep learning and language modeling (SciBERT).
Event hosted by Drive Sweden and AI Sweden, with guests from the University of Borås Data as Impact Lab.