Building End-To-End Dialogue Systems Using Generative Hierarchical Neural Network Models

Serban, Iulian Vlad; Sordoni, Alessandro; Bengio, Yoshua; Courville, Aaron; Pineau, Joëlle

doi:10.1609/aaai.v30i1.9883

Cited by 1,089 publications

(244 citation statements)

References 26 publications

Supporting

Mentioning

242

Contrasting

Unclassified

Order By: Relevance

“…However, these studies, mostly inspired by psychology findings, are either rule-based or limited to small-scale data. Recently, neural models trained on large-scale data have advanced open-domain conversation generation significantly (Ritter, Cherry, and Dolan 2011;Vinyals and Le 2015;Shang, Lu, and Li 2015;Serban et al 2016). Most of these models aim to improve the content quality of conversation generation (Gu et al 2016;Li et al 2016a;Xing et al 2017;Mou et al 2016;Li et al 2016b).…”

Section: Introductionmentioning

confidence: 99%

Emotional Chatting Machine: Emotional Conversation Generation with Internal and External Memory

Zhou

Huang

Zhang

et al. 2018

AAAI

482

113

View full text Add to dashboard Cite

Perception and expression of emotion are key factors to the success of dialogue systems or conversational agents. However, this problem has not been studied in large-scale conversation generation so far. In this paper, we propose Emotional Chatting Machine (ECM) that can generate appropriate responses not only in content (relevant and grammatical) but also in emotion (emotionally consistent). To the best of our knowledge, this is the first work that addresses the emotion factor in large-scale conversation generation. ECM addresses the factor using three new mechanisms that respectively (1) models the high-level abstraction of emotion expressions by embedding emotion categories, (2) captures the change of implicit internal emotion states, and (3) uses explicit emotion expressions with an external emotion vocabulary. Experiments show that the proposed model can generate responses appropriate not only in content but also in emotion.

show abstract

Section: Introductionmentioning

confidence: 99%

Emotional Chatting Machine: Emotional Conversation Generation with Internal and External Memory

Zhou

Huang

Zhang

et al. 2018

AAAI

482

113

View full text Add to dashboard Cite

show abstract

“…Data-driven conversational models generally fall into two categories: retrieval-based methods (Lowe et al 2015b;2016a;Zhou et al 2016), which select a response from a predefined repository, and generation-based methods (Ritter, Cherry, and Dolan 2011;Serban et al 2016;Vinyals and Le 2015), which employ an encoder-decoder framework where the message is encoded into a vector representation and, then, fed to the decoder to generate the response. The latter is more natural (as it does not require a response repository) yet suffers from generating dull or vague responses and generally needs a great amount of training data.…”

Section: Related Work Conversational Modelsmentioning

confidence: 99%

Augmenting End-to-End Dialogue Systems With Commonsense Knowledge

Young

Cambria

Chaturvedi

et al. 2018

AAAI

210

View full text Add to dashboard Cite

Building dialogue systems that can converse naturally with humans is a challenging yet intriguing problem of artificial intelligence. In open-domain human-computer conversation, where the conversational agent is expected to respond to human utterances in an interesting and engaging way, commonsense knowledge has to be integrated into the model effectively. In this paper, we investigate the impact of providing commonsense knowledge about the concepts covered in the dialogue. Our model represents the first attempt to integrating a large commonsense knowledge base into end-to-end conversational models. In the retrieval-based scenario, we propose a model to jointly take into account message content and related commonsense for selecting an appropriate response. Our experiments suggest that the knowledge-augmented models are superior to their knowledge-free counterparts.

show abstract

“…All the proposed models were implemented during the testing, following Li and Jurafsky (2016). We used the bi-directional recurrent neural network with gated recurrent units (Bi-GRU RNN) (Serban et al 2016a) to capture the information along the word sequences. To train the neural conversation models, we followed the hyperparameter settings in (Shang, Lu, and Li 2015;Song et al 2016).…”

Section: Experiments Experimental Setupsmentioning

confidence: 99%

Towards a Neural Conversation Model With Diversity Net Using Determinantal Point Processes

Song

Yan

Feng

et al. 2018

AAAI

View full text Add to dashboard Cite

Typically, neural conversation systems generate replies based on the sequence-to-sequence (seq2seq) model. seq2seq tends to produce safe and universal replies, which suffers from the lack of diversity and information. Determinantal Point Processes (DPPs) is a probabilistic model defined on item sets, which can select the items with good diversity and quality. In this paper, we investigate the diversity issue in two different aspects, namely query-level and system-level diversity. We propose a novel framework which organically combines seq2seq model with Determinantal Point Processes (DPPs). The new framework achieves high quality in generated reply and significantly improves the diversity among them. Experiments show that our model achieves the best performance among various baselines in terms of both quality and diversity.

show abstract

Building End-To-End Dialogue Systems Using Generative Hierarchical Neural Network Models

Cited by 1,089 publications

References 26 publications

Emotional Chatting Machine: Emotional Conversation Generation with Internal and External Memory

Emotional Chatting Machine: Emotional Conversation Generation with Internal and External Memory

Augmenting End-to-End Dialogue Systems With Commonsense Knowledge

Towards a Neural Conversation Model With Diversity Net Using Determinantal Point Processes

Contact Info

Product

Resources

About