| Title | : | Incorporating External Knowledge in Domain Specific Conversation Systems |
| Speaker | : | Nikita Moghe (IITM) |
| Details | : | Tue, 16 Apr, 2019 2:00 PM @ A M Turing Hall |
| Abstract: | : | Existing dialog datasets contain a sequence of utterances and responses without any explicit background knowledge associated with them. This has resulted in the development of models which treat conversation as a sequence-to-sequence generation task (i.e., given a sequence of utterances generate the response sequence). This is not only an overly simplistic view of conversation but it is also emphatically different from the way humans converse by heavily relying on their background knowledge about the topic. The aim of this work, thus, is to facilitate the development of such natural conversation models which mimic the human process of conversing using appropriate background knowledge. With this view, we create a new dataset containing movie chats wherein each response is explicitly generated by copying and/or modifying sentences from unstructured background knowledge such as plots, comments and reviews about the movie. We establish baseline results on this dataset using three different paradigms (i) pure generation based models which ignore the background knowledge (ii) generation based models which learn to copy information from the background knowledge when required and (iii) span prediction based models which predict the appropriate response span in the background knowledge. We further propose improvements to the generation with copy paradigm using structural information. Specifically, we improve upon the response generation task by investigating combinations of the Graph Convolutional Networks (GCN) and Recurrent Neural Networks(RNN) and different kinds of structural information. We observe that using co-reference, entity, and dependency structures together as structural information with a specific GCN-RNN combination leads to state-of-the art performance on this task. |
