Boosting Question Answering by Deep Entity Recognition
This work addresses the challenge of entity recognition in question answering for the Polish language, offering a novel method to improve accuracy for questions beyond standard named entities, though it is incremental as it builds on existing NER approaches.
The authors tackled the problem of open-domain factoid question answering in Polish by developing a system that extracts precise entity strings, introducing Deep Entity Recognition (DeepER) to handle entity references beyond traditional named entities. The system was evaluated on over a thousand questions from a quiz TV show using Polish Wikipedia, with manual evaluation showing improved ability to answer questions requiring non-traditional entity categories.
In this paper an open-domain factoid question answering system for Polish, RAFAEL, is presented. The system goes beyond finding an answering sentence; it also extracts a single string, corresponding to the required entity. Herein the focus is placed on different approaches to entity recognition, essential for retrieving information matching question constraints. Apart from traditional approach, including named entity recognition (NER) solutions, a novel technique, called Deep Entity Recognition (DeepER), is introduced and implemented. It allows a comprehensive search of all forms of entity references matching a given WordNet synset (e.g. an impressionist), based on a previously assembled entity library. It has been created by analysing the first sentences of encyclopaedia entries and disambiguation and redirect pages. DeepER also provides automatic evaluation, which makes possible numerous experiments, including over a thousand questions from a quiz TV show answered on the grounds of Polish Wikipedia. The final results of a manual evaluation on a separate question set show that the strength of DeepER approach lies in its ability to answer questions that demand answers beyond the traditional categories of named entities.