AskNow - Question Answering

A core part of decision making processes involves the meticulous collection of relevant knowledge. This involves hunting down information about facts and events from the web, as well as querying data from within the organization. At Fraunhofer IAIS we are working at the cutting edge of a rapidly developing research field: the difficulty of making information access intuitive.
 
Question Answering provides an alternative, more intuitive method for collecting knowledge over various sources, and by forgoing query language expertise, organizations can now focus on what matters.

Our Solution:

We develop systems capable of answering questions over the knowledge that organizations generate and have accumulated. With its assistance, enterprises can forgo convoluted queries and simply gather information like they would from a human being, in plain natural language. To accomplish this feat, our experts have employed innovative concepts and technologies such as »Linked Data«, »Deep Learning« and »Natural Language Processing«.

As a result, different industries can benefit from using our system, for example, it can be used as a supportive module for customer service in telecommunication companies, or as an integrated part of the in-car systems in the automotive industry, and likewise to countless public sectors portals.

Interview »Question Answering«

With the click on the play button an external video from wwww.youtube.com is loaded and started. Your data is possible transferred and stored to third party. Do not start the video if you disagree.

Prof. Jens Lehmann explains the concept of Question Answering, what are the potential applications of this technology, and how we will interact with computers in the future.  

Our Methods, Technologies and Expertise:

  • Hybrid Semantic Parsing
  • Deep Learning-based Question Answering
  • Knowledge Graphs
  • Ontology Modelling

Applications and projects

HOBBIT aims at abolishing the barriers in the adoption and deployment of Big Linked Data by European companies, by means of open benchmarking reports that allow them to assess the fitness of existing solutions for their purposes.  

One evaulation campaign is concerned with question-answering systems based on linked data and featuring multilinguality. 

Project duration: December 2015 - November 2018