Generating Long and Diverse Responses with Neural Conversation Models

By Louis Yuanlong Shao, Stephan Gouws, Denny Britz, Anna Goldie, Brian Strope, R. Kurzweil. 2017

65 citations recorded by the source corpus.

Subfields: Natural Language Processing, Generative Models

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Natural Language Processing

In Generative Models

Citations in the corpus

Hierarchical Neural Story Generation (2018, methodology): “We ﬁnd this sampling strategy substantially more effective than beam search, which tends to produce common phrases and repetitive text from the training set (Vijayaku-mar et al., 2016; Shao et al., 2017 Table 3: Perplexity on W RITING P ROMPTS .”
Adversarial Learning for Neural Dialogue Generation (2017, methodology): “This resembles the N-choose-1 metric described in Shao et al. (2017).”
A Simple, Fast Diverse Decoding Algorithm for Neural Generation (2016, methodology): “In neural response generation, a re-ranking step helps avoid generating dull or generic responses (Li et al., 2015a; Sordoni et al., 2015; Shao et al., 2016).”
A Survey on Dialogue Systems: Recent Advances and New Frontiers (2017, background): “[62] introduced a stochastic beam-search procedure, while [34] added an additional term for beam search scoring to penalize the siblings–expansions of the same parent node in the search.”
Commonsense Knowledge Aware Conversation Generation with Graph Attention (2018, methodology): “These models developed various techniques to improve the content quality of generated responses, including diversity promotion [Li et al. , 2016; Shao et al. , 2017], considering additional information [Xing et al. , 2017; Mou et al. , 2016], and handling unknown words [Gu et al. , 2016 ] .”
Steering Output Style and Topic in Neural Response Generation (2017, background): “Decoding only according to this function will thus result in only low-frequency words and un-grammatical sentences, behavior also noted by (Li et al., 2015; Shao et al., 2017).”
Recent advances in deep learning based dialogue systems: a systematic survey (2021, background): “Some works further added a reranking stage to select more diverse responses in the generated N-best list (Li et al., 2015; Sordoni et al., 2015b; Shao et al., 2017).”
Extractive Summarization of Long Documents by Combining Global and Local Context (2019, background): “However when it comes to long documents, these models tend to struggle with longer sequences because at each decoding step, the de-coder needs to learn to construct a context vector capturing relevant information from all the tokens in the source sequence (Shao et al., 2017).”
Better Conversations by Modeling, Filtering, and Optimizing for Coherence and Diversity (2018, methodology): “Many works address the lack of diversity and coherence in E-D outputs (Sountsov and Sarawagi, 2016; Wei et al., 2017) 2017) in addition to using a self-attention model.”
Cluster-based beam search for pointer-generator chatbot grounded by knowledge (2020, background): “(Shao et al. 2016) introduced a stochastic beam-search algorithm with segmentby-segment reranking to inject diversity earlier in the decoding process.”