Daily Reading

A Diversity-Promoting Objective Function for Neural Conversation Models

Summary & Intuitions

• mutual information between source (message) and target (response)
• lack of theoretical guarantee

Contributions

• decompose formula of mutual information:
  • anti-lm: penalize not only high-frequency, generic responses but also grammatical sentence $\rightarrow$ weights of tokens decrease monotonically (early important + lm dominant later)
  • bidi-lm: not searching but reranking (generate grammatical sequences and then re-rank them according to the objective of inversed probability)

ClarQ: A large-scale and diverse dataset for Clarification Question Generation

Summary & Intuitions

• collected post and corresponding comments
• improve precision then recall

Contributions

• down-sampling: confident samples as training set of next iteration
• up-sampling: classifier from $t$ iteration classify samples from $t-1$