Document worth reading: “A Survey of Natural Language Generation Techniques with a Focus on Dialogue Systems – Past, Present and Future Directions”
One of the hardest points inside the house of Natural Language Processing and Artificial Intelligence is routinely producing language that is coherent and understandable to folks. Teaching machines how one can converse as folks do falls beneath the broad umbrella of Natural Language Generation. Recent years have seen unprecedented growth inside the amount of evaluation articles printed on this subject in conferences and journals every by tutorial and enterprise researchers. There have moreover been a quantity of workshops organized alongside top-tier NLP conferences devoted significantly to this disadvantage. All this train makes it laborious to clearly define the state of the sphere and trigger about its future directions. In this work, we provide a top level view of this important and thriving house, masking typical approaches, statistical approaches and moreover approaches that use deep neural networks. We current a full analysis within the path of developing open space dialogue strategies, a vital utility of pure language period. We uncover that, predominantly, the approaches for developing dialogue strategies use seq2seq or language fashions construction. Notably, we set up three important areas of extra evaluation within the path of developing extra sensible dialogue strategies: 1) incorporating greater context, collectively with dialog context and world information; 2) together with personae or character inside the NLG system; and 3) overcoming uninteresting and generic responses that have an impact on the usual of system-produced responses. We current pointers on how one can deal with these open points through the use of cognitive architectures that mimic human language understanding and period capabilities. A Survey of Natural Language Generation Techniques with a Focus on Dialogue Systems – Past, Present and Future Directions