Answering with Cases: A CBR Approach to Deep Learning Kareem Amin 1,3 , Stelios Kapetanakis 4,5 , Klaus-Dieter Althoff 1,2 , Andreas Dengel 1,3 , and Miltos Petridis 6 1 German Research Center for Artificial Intelligence, Smart Data and Knowledge Services, Trippstadter Strae 122, 67663 Kaiserslautern, Germany, kareem.amin,klaus-dieter.althoff,andreas.dengel@dfki.uni-kl.de 2 Institute of Computer Science, Intelligent Information Systems Lab, University of Hildesheim, Hildesheim, Germany, 3 Kaiserslautern University, P.O. Box 3049, 67663 Kaiserslautern, Germany 4 School of Computing Engineering and Mathematics, University of Brighton, s.kapetanakis@brighton.ac.uk 5 Gluru Research, Gluru, London stelios@gluru.co 6 Department of Computing, University of Middlesex, London, UK m.petridis@mdx.ac.uk Abstract. Every year tenths of thousands of customer support engi- neers around the world deal with, and proactively solve, complex help- desk tickets. Daily, almost every customer support expert will turn his/her attention to a prioritization strategy, to achieve the best possible result. To assist with this, in this paper we describe a novel case-based reasoning application to address the tasks of: high solution accuracy and shorter prediction resolution time. We describe how appropriate cases can be generated to assist engineers and how our solution can scale over time to produce domain-specific reusable cases for similar problems. Our work is evaluated using data from 5000 cases from the automotive industry. Keywords: Case-based Reasoning, Deep Learning, Natural Language Pro- cessing 1 Introduction Effective Customer Support can be a challenge. Both for a company and for a trained system engineer it depends on endless hours of case scanning, a large va- riety of complex factors and in cases obscure case definitions.To complete a series of tickets successfully a help-desk engineer needs an appropriate prioritization strategy for every working day. The engineer must select a suitable prioritiza- tion route, based on the problem description, complexity and historical evidence upon its possible solution. The aim of this work is to help support engineers to achieve the best possible outcome for a given ticket. We propose case-based