9:00 AM - 10:40 AM
[4Pin1-11] Divide and Generate: Neural Generation of Complex Sentences
Keywords:sentence generation, dialogue system, evaluation
In recent years, research on chat dialogue systems has been attracted much attention. Typical chat dialogue system selects and outputs an appropriate response from a dialogue database for the input user utterance. However, it is not possible to output an appropriate response if the coverage of the database is limited. Therefore it is necessary to augment the database beforehand.
In this research, in order to amplify various kinds of responses in the database, we propose a task to generate a complex sentence from a simple sentence. We first divide a complex sentence into a main clause and a subordinate clause to learn a generator model of modifiers, and then use the model to generate a modifier clause to create a complex sentence from a simple sentence. We present an automatic evaluation metric to estimate the quality of models.
In this research, in order to amplify various kinds of responses in the database, we propose a task to generate a complex sentence from a simple sentence. We first divide a complex sentence into a main clause and a subordinate clause to learn a generator model of modifiers, and then use the model to generate a modifier clause to create a complex sentence from a simple sentence. We present an automatic evaluation metric to estimate the quality of models.