An Overview of Utilizing Knowledge Bases in Neural Networks for Question Answering

The Power and Potentials of Flexible Query Answering Systems: A Critical and Comprehensive Analysis

“…A potential alternative approach is exploiting multi-modality in which sequences of images, audio files and videos, and text illustrating or exemplifying the requested procedures are retrieved by leveraging on neural systems' ability of modelling multi-modal information sources [126]. In Visual Query Answering (VQA) [127] a NL answer is yielded by the system as a result of evaluating a NL question asking for information on an image content provided as input to the system.…”

Heuristic Acquisition for Data Science

“…The seventh paper (Kafle et al 2020) deals with the important problem of knowledge representation in knowledge bases for assisting neural networks in understanding natural language. It used to be held that recurrent neural networks (RNNs)could learn to understand natural language; but this approach has recently yielded to model-based approaches.…”