A Taxonomy for Deep Learning in Natural Language Processing

Despite a large number of available techniques around Deep Learning in Natural Language Processing (NLP), no holistic framework exists which supports researchers and practitioners to organise knowledge when designing, comparing evaluating NLP applications. This paper addresses this lack by developing taxonomy for Processing. Based on systematic literature review as proposed Webster Watson vom Brocke et al. the iterative development process Nickerson we derived five novel dimensions 38 characteristics based sample 205 papers. Our research suggests, that approach can be distinguished were partly from CRISP-DM methodology: application understanding, data preparation, modeling, learning technique evaluation. We, therefore, hope provide guidance support using design, compare evaluate