Pie & AI Suisse - Question Answering & recent NLP Breakthroughs

Our 2nd "Pie & AI Suisse" Meetup will be all about how to answer questions as a machine and the latest research findings ---

Question answering on unstructured text is a challenging Natural Language Processing (NLP) task in which all facets of human language need to be considered to correctly identify the correct answer from any arbitrary text.

We utilize the question answering use case as an example to show how challenging it is to model and implement a computer system for solving such tasks. We review recent NLP breakthroughs, which together with an open source culture are setting new standards for automation. As a result the time to market, from research to production, is getting faster and faster.

We also show a live demo and how you can create such a questions answering system of an arbitrary website in only ~20 lines of code.

BIO:

Arlind Nocaj is a computer science enthusiast and architect who likes to bring the "aha effect" to people by using the power of data and algorithms. With a background in algorithmics as well as network analysis and visualization, his current areas of interest are data analytics, machine learning and distributed systems. He brings innovation to clients by combining data and algorithms to automate use cases end to end.
Before joining the Emerging and Growth team in Accenture, he was working as a researcher at the University of Konstanz and as research developer at Microsoft Research in Redmond.