Goal-oriented dialog has been given attention due to its numerous applications in artificial intelligence. Goal-oriented dialogue tasks occur when a questioner asks an action-oriented question and an answerer responds with the intent of letting the questioner know a correct action to take. To ask the adequate question, deep learning and reinforcement learning have been recently applied. However, these approaches struggle to find a competent recurrent neural questioner, owing to the complexity of learning a series of sentences. Motivated by theory of mind, we propose Answerer in Questioner's Mind (AQM), a novel information theoretic algorithm for goal-oriented dialog. With AQM, a questioner asks and infers based on an approximated probabilistic model of the answerer.

There are words with multiple clusters (Figure 1b).

Neural Information Processing Systems http://nips.cc/

Personalization usually gets a big plus in many contexts. Think about many potential axes, including language, geographic location, task orientation, product/service description, medical condition, garment size, food allergies, educational focus, job category, news preference: The list is long. The consequences of this kind of personalization are usually seen as useful because the system is intended to produce results tailored to an individual's interests. In the advertising world, this is often extremely valuable since the information is targeted at a specific need or interest. The same can be said for many other specific cases in which a need or interest is satisfied more effectively.