HPI in-memory-based Database System in Task 2b of BioASQ.
Mariana Neves · CLEF (Working Notes) · 2014
We describe the participation of the Hasso-Plattner Institut (HPI) team in the BioASQ challenge and in particular in the Task 2b (Phase A), which consisted of providing results for potentially relevant concepts, documents and passages, given a certain question. Our systems relies on the in-memory based database (IMDB) technology and built-in text analysis provided by a IMDB database, as well as external resources, such as BioPortal and the pre-defined ontologies and terminologies to be used for the task. We present an evaluation of a preliminary version of our system on the training dataset (310 questions) and on the three of the test batches of 100 questions. Our results were particularly good for the passage retrieval, including a first position in one of the batches, which prove the feasibility of our approach for the question answering task.