Emotional Conversation Generation Research Articles

An Emotional Dialogue System Using Conditional Generative Adversarial Networks with a Sequence-to-Sequence Transformer Encoder

Understanding the expression of emotion and generating appropriate responses are key steps toward constructing emotional, conversational agents. In this paper, we propose a framework for single-turn emotional conversation generation, and there are three main components in our model, namely, a sequence-to-sequence model with stacked encoders, a conditional variational autoencoder, and conditional generative adversarial networks. For the sequence-to-sequence model with stacked encoders, we designed a two-layer encoder by combining Transformer with gated recurrent units-based neural networks. Because of the flexibility of the sequence-to-sequence model, we adopted a conditional variational autoencoder in our framework, which uses latent variables to learn a distribution over potential responses and generates diverse responses. Furthermore, we regard a conditional variational autoencoder-based, sequence-to-sequence model as the generative model, and the training of the generative model is assisted by both a content discriminator and an emotion classifier, which assists our model in promoting content information and emotion expression. We use automated evaluation and human evaluation to evaluate our model and baselines on the NTCIR short text conversation task (STC-3) Chinese emotional conversation generation (CECG) Subtask dataset [44], and the experimental results demonstrate that our proposed framework can generate semantically reasonable and emotionally appropriate responses.

Leveraging hierarchical semantic‐emotional memory in emotional conversation generation

AbstractHandling emotions in human‐computer dialogues has emerged as a challenging task which requires artificial intelligence systems to generate emotional responses by jointly perceiving the emotion involved in the input posts and incorporating it into the generation of semantically coherent and emotionally reasonable responses. However, most previous works generate emotional responses solely from input posts, which do not take full advantage of the training corpus and suffer from generating generic responses. In this study, we introduce a hierarchical semantic‐emotional memory module for emotional conversation generation (called HSEMEC), which can learn abstract semantic conversation patterns and emotional information from the large training corpus. The learnt semantic and emotional knowledge helps to enrich the post representation and assist the emotional conversation generation. Comprehensive experiments on a large real‐world conversation corpus show that HSEMEC can outperform the strong baselines on both automatic and manual evaluation. For reproducibility, we release the code and data publicly at: https://github.com/siat‐nlp/HSEMEC‐code‐data.

Enhancing Integrity Modeling for Emotional Conversation Generation

Establishing an open-domain emotional conversation generation with content and emotional dependence is a key step toward intelligent interactions. However, the researchers only dedicated to adding the thematic content or emotional words and handling both factors are not yet properly solved. In this article, we aim to address the issue of integrity in emotional content which indicates the response has a dependency relationship between theme and emotion. We propose a thematic–emotional conversation model, which consists of parallel channels and joint encoding, to maximize integrity in conversation generation. We first use the encoding of the parallel channel to strengthen the decoding weight of the theme and emotion. The joint encoding of the dual attention mechanism and the thematic emotional tradeoff strategy are involved in the encoding to ensure that the theme complements the emotional content during the response generation. After that, the bidirectional training interactions between post and response are calculated to derive the new response inference vectors. Experiment results suggest that the proposed model is effective not only in content integrity but also in emotion.

Emotional conversation generation with heterogeneous graph neural network

The successful emotional conversation system depends on sufficient perception and appropriate expression of emotions. In a real-life conversation, humans firstly instinctively perceive emotions from multi-source information, including the emotion flow hidden in dialogue history, facial expressions, audio, and personalities of speakers. Then, they convey suitable emotions according to their personalities, but these multiple types of information are insufficiently exploited in emotional conversation fields. To address this issue, in this paper, we propose a heterogeneous graph-based model for emotional conversation generation. Firstly, we design a Heterogeneous Graph-Based Encoder to represent the conversation content (i.e., the dialogue history, its emotion flow, facial expressions, audio, and speakers' personalities) with a heterogeneous graph neural network, and then predict suitable emotions for feedback. Secondly, we employ an Emotion-Personality-Aware Decoder to generate a response relevant to the conversation context as well as with appropriate emotions, through taking the encoded graph representations, the predicted emotions by the encoder and the personality of the current speaker as inputs. Experiments on both automatic and human evaluation show that our method can effectively perceive emotions from multi-source knowledge and generate a satisfactory response. Furthermore, based on the up-to-date text generator BART, our model still can achieve consistent improvement, which significantly outperforms some existing state-of-the-art models.

Informative and diverse emotional conversation generation with variational recurrent pointer-generator

Informative and diverse emotional conversation generation with variational recurrent pointer-generator

Emotional Conversation Generation Orientated Syntactically Constrained Bidirectional-Asynchronous Framework

The field of open-domain conversation generation using deep neural networks has attracted increasing attention from researchers for several years. However, traditional neural language models tend to generate safe, generic reply with poor logic and no emotion. In this paper, an emotional conversation generation orientated syntactically constrained bidirectional-asynchronous framework called E-SCBA is proposed to generate meaningful (logical and emotional) reply. In E-SCBA, pre-generated emotion keyword and topic keyword are asynchronously introduced into the reply during the generation, and the process of decoding is much different from the most existing methods that generates reply from the first word to the end. A newly designed bidirectional-asynchronous decoder with the multi-stage strategy is proposed to support this idea, which ensures the fluency and grammaticality of reply by making full use of syntactic constraint. Through the experiments, the results show that our framework not only improves the diversity of replies, but gains a boost on both logic and emotion compared with baselines as well.

Infusing Multi-Source Knowledge with Heterogeneous Graph Neural Network for Emotional Conversation Generation

The success of emotional conversation systems depends on sufficient perception and appropriate expression of emotions. In a real-world conversation, we firstly instinctively perceive emotions from multi-source information, including the emotion flow of dialogue history, facial expressions, and personalities of speakers, and then express suitable emotions according to our personalities, but these multiple types of information are insufficiently exploited in emotional conversation fields. To address this issue, we propose a heterogeneous graph-based model for emotional conversation generation. Specifically, we design a Heterogeneous Graph-Based Encoder to represent the conversation content (i.e., the dialogue history, its emotion flow, facial expressions, and speakers' personalities) with a heterogeneous graph neural network, and then predict suitable emotions for feedback. After that, we employ an Emotion-Personality-Aware Decoder to generate a response not only relevant to the conversation context but also with appropriate emotions, by taking the encoded graph representations, the predicted emotions from the encoder and the personality of the current speaker as inputs. Experimental results show that our model can effectively perceive emotions from multi-source knowledge and generate a satisfactory response, which significantly outperforms previous state-of-the-art models.

Predicting Emotion Reactions for Human–Computer Conversation: A Variational Approach

Although emotional conversation generation has attracted widespread attention in recent years, research works on the emotional interactional mechanism are still scarce, which makes it difficult for existing emotional dialogue system to automatically determine a suitable emotion type for conversation generation. Such a response emotion planning task is often difficult due to the “gap” problem: we need to predict the emotional probability distribution of the upcoming responses, which have not yet been generated. In this article, we propose a novel method, namely interactional emotion learning (IEL), which adopts an intuitive way to eliminate the “gap” problem: we design a variational learning network, called potential response learning, to learn the latent distribution of responses for given conversation in a semantic space. Then, we predict an appropriate emotion for response generation based on both the dialogue context and the learned latent distribution. Extensive experiments have been performed on three off-the-shelf conversation datasets, and the experimental results show that the proposed variational learning network significantly boosts the prediction ability of our approach, and our IEL method outperforms the state-of-the-art dialogue classification methods in the emotion planning task.

Topic-extended Emotional Conversation Generation Model Based on Joint Decoding

The research on the expression of emotion in human-computer dialogue can greatly improve the user experience. Existing research has paid a lot of attention to how to generate specific emotional content and how to improve the extraction rate of emotions, while ignoring the reduction of emotion expression caused by factors such as topics and emotions added to the encoder. This paper proposes a novel Topic-extended Emotional Conversation Generation Model Based on Joint Decoding (TECM-JD). The model embeds the specified emotion category as an additional input into the emotional independent unit of the decoder, in order to reduce the expression of the content affected by adding emotion into the model. The joint attention mechanism is used to obtain the input sequence content and the input sequence topic word content obtained by the Twitter LDA model, which ensures that the output topic and the input are under the same topic. The experimental results show that the proposed model can generate richer emotional content related to the topic and have good performance and are superior to traditional dialogue models.

Non-deterministic and emotional chatting machine: learning emotional conversation generation using conditional variational autoencoders

Conversational responses are non-trivial for artificial conversational agents. Artificial responses should not only be meaningful and plausible, but should also (1) have an emotional context and (2) should be non-deterministic (i.e., vary given the same input). The two factors enumerated, respectively, above are involved and this is demonstrated such that previous studies have tackled them individually. This paper is the first to tackle them together. Specifically, we present two models both based upon conditional variational autoencoders. The first model learns disentangled latent representations to generate conversational responses given a specific emotion. The other model explicitly learns different emotions using a mixture of multivariate Gaussian distributions. Experiments show that our proposed models can generate more plausible and diverse conversation responses in accordance with designated emotions compared to baseline approaches.

A Control Unit for Emotional Conversation Generation

Emotional conversation generation model predicts the response according to the current words and the emotional words. However, the researchers only dedicated to adding more emotional words in the conversation generation model to retain the taste of chat users without considering whether the emotion of a response is suitable for human conversations or not. In this paper, we aim to address the issue of emotion drift which indicates the emotion of a response is not the same category as its post in human conversations. We propose a control unit framework, which consists of emotional channels and word-level attention mechanism, to incorporate natural and smooth emotional words into conversation generation. Emotional channel consists six channels, namely like, sadness, disgust, anger, happiness and other ones, which provides strategy choice control unit to generate emotional words. To improve the importance of emotional content, we use the word-level attention mechanism in emotional channel for acquiring a better emotional decoding response. Experimental results suggest that the proposed model is effective not only in generate content but also in emotion.

Emotional Conversation Generation Based on a Bayesian Deep Neural Network

The field of conversation generation using neural networks has attracted increasing attention from researchers for several years. However, traditional neural language models tend to generate a generic reply with poor semantic logic and no emotion. This article proposes an emotional conversation generation model based on a Bayesian deep neural network that can generate replies with rich emotions, clear themes, and diverse sentences. The topic and emotional keywords of the replies are pregenerated by introducing commonsense knowledge in the model. The reply is divided into multiple clauses, and then a multidimensional generator based on the transformer mechanism proposed in this article is used to iteratively generate clauses from two dimensions: sentence granularity and sentence structure. Subjective and objective experiments prove that compared with existing models, the proposed model effectively improves the semantic logic and emotional accuracy of replies. This model also significantly enhances the diversity of replies, largely overcoming the shortcomings of traditional models that generate safe replies.

Topic-enhanced emotional conversation generation with attention mechanism

Emotional conversation generation has elicited a wide interest in both academia and industry. However, existing emotional neural conversation systems tend to ignore the necessity to combine topic and emotion in generating responses, possibly leading to a decline in the quality of responses. This paper proposes a topic-enhanced emotional conversation generation model that incorporates emotional factors and topic information into the conversation system, by using two mechanisms. First, we use a Twitter latent Dirichlet allocation (LDA) model to obtain topic words of the input sequences as extra prior information, ensuring the consistency of content between posts and responses for emotional conversation generation. Second, the system uses a dynamic emotional attention mechanism to adaptively acquire content-related and affective information of the input texts and extra topics. The advantage of this study lies in the fact that the presented model can generate abundant emotional responses, with the contents being related and diverse. To demonstrate the effectiveness of our method, we conduct extensive experiments on large-scale Weibo post–response pairs. Experimental results show that our method achieves good performance, even outperforming some existing models.

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

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.