survey-in-dialog-system

This repo includes two comprehensive surveys of recent advances and trends in technology of data-driven dialog system.

2017_survey

Chen, Hongshen, et al. "A Survey on Dialogue Systems: Recent Advances and New Frontiers." arXiv preprint arXiv:1711.01731 (2017).

This survey contains two topics: 1.Goal-oriented dialog system; 2.Open domain dialog system. For topic 1, it includes sub-topics: Pipeline methods(NLU -> DST-> policy learning ->NLG) and End-to-End methods. For topic 2, it includes sub-topics: Generative-based Model (seq2seq model, context-sensitive model, response diversity, response with topic and personality, incoporate external KB and Reinforcement learning), Retrieval-based model (Single-turn Response Matching and Multi-turn Response Matching) and Hybrid models which combine generative based and retrieval based models.

corpora_survey

Serban, Iulian Vlad, et al. "A survey of available corpora for building data-driven dialogue systems." arXiv preprint arXiv:1512.05742 (2015).

This survey contains all kinds of corpus used to build dialog system.