Brain-Inspired Approaches to Natural Language Processing and Explainable Artificial Intelligence

Abstract

Like no other medium, the World Wide Web became the major information source for many people within the last years, some even call it the brain of mankind. For any arising questions, any facts needed, or any multimedia content wanted, a web page providing the respective information seems to exist. Likewise, it seems that sometimes there is nothing that has not been thought, written, painted, or expressed in any other form before: most users simply feel overwhelmed by the flood of available information. Consequently, there is a need for new technologies for autonomous self-management, more timely information handling, processing, and the user’s interaction with such huge amounts of data. Indeed, Einstein’s saying Look deep into nature, then you will understand everything better is a big inspiration and challenge to find the required, new solutions. At this point, a short overview is given of existing organizational and functional principles, which have been derived from nature and in particular the human brain and which could be adapted to realize the desired, new methods for natural language processing. The methods mostly follow the strict natural design principle of locality, i.e. work without overseeing the whole system or full set of data, and exhibit a high degree of parallelism. Also, specific application fields for them will be discussed.