Dialogue Systems Research Articles

DialGNN: Heterogeneous Graph Neural Networks for Dialogue Classification

Dialogue systems have attracted growing research interests due to its widespread applications in various domains. However, most research work focus on sentence-level intent recognition to interpret user utterances in dialogue systems, while the comprehension of the whole documents has not attracted sufficient attention. In this paper, we propose DialGNN, a heterogeneous graph neural network framework tailored for the problem of dialogue classification which takes the entire dialogue as input. Specifically, a heterogeneous graph is constructed with nodes in different levels of semantic granularity. The graph framework allows flexible integration of various pre-trained language representation models, such as BERT and its variants, which endows DialGNN with powerful text representational capabilities. DialGNN outperforms on CM and ECS datasets, which demonstrates robustness and the effectiveness. Specifically, our model achieves a notable enhancement in performance, optimizing the classification of document-level dialogue text. The implementation of DialGNN and related data are shared through https://github.com/821code/DialGNN.

Revisiting clustering for efficient unsupervised dialogue structure induction

In the development of a task-oriented dialogue system, defining the dialogue structure is a time-consuming task. Hence, several works have looked into automatically inferring it from data, e.g., actual conversations between a customer and a support agent. To recover such dialogue structure, recent methods based on discrete variational models learn to jointly encode and cluster utterances in dialogue states, but (i) represent utterances by only considering preceding dialogue context, and (ii) are slow to train since they are optimized with a compute-expensive decoding objective. We revisit and improve upon an existing efficient pipeline approach, commonly adopted as a baseline, that first encodes utterances and then clusters them with k-means to induce the dialogue structure. However, the existing approach represents utterances as bag-of-words or skip-thought vectors, which have been shown to perform poorly in semantic similarity tasks, and without considering dialogue context. We therefore first investigate the use of more powerful transformer-based encoders for encoding utterances. Next, we propose ellodar, a method for learning representations that capture both preceding and subsequent dialogue context, inspired by word-to-vec training strategies. ellodar is efficient since representations are learned directly in the encoding space by finetuning just a single linear layer on top of a frozen sentence encoder with a vector-to-vector regression training objective. Extensive experiments on representative datasets for dialogue structure induction (SimDial, Schema Guided Dialogues, DSTC2, and CamRest676) demonstrate that in terms of effectiveness to induce the correct dialogue structure, (i) clustering utterances represented by transformed-based encoders improves recent joint models by 13%–32% on standard cluster metrics, and (ii) clustering ellodar’s representations yields additional improvements ranging from +20% to +26%, with speedups of ×\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\ imes $$\\end{document} 10\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\ extbf{10}$$\\end{document}–104\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\ extbf{10}^{\ extbf{4}}$$\\end{document} compared to the recent joint models.

Development of an intelligent assistant for selection of goods in the process of dialogue with the user

This article is devoted to the development of methods for creating intelligent assistants. Intelligent assistants can be used in call centers to solve customer problems, to solve technical support tasks, to help people with disabilities, to help in choosing goods, etc. We consider intelligent assistants that engage in argumentative dialogue with users, aimed at finding goods and services that maximally satisfy users’ wants and needs. The development of the intelligent assistant is based on a four-level model of the subject domain and a semantic model of the user. The system under development automates the process of search and decision justification through the reuse of domain cases: accumulated knowledge about previous dialogues with users. This gives the system we developed an advantage over existing analogues, which are incapable of reusing knowledge about previous dialogues. The paper develops a case-based approach to building an intelligent system capable of reasoning about its responses. For this purpose, an argumentation graph is constructed, methods for structuring domain cases are developed, and ontological homomorphisms are used to transform the available domain cases into a finished solution. A description of model-theoretical methods for constructing intelligent assistants is presented. The cases of goods, users and dialogues of an intelligent assistant with users are formally described in the form of partial models. The transformation of domain cases and similarity of cases are formalized using ontological homomorphisms of partial models. The purpose of the developed dialogue system is not only to select a solution according to the user’s request, but also to find out the tasks that the user is going to solve, to analyze his argumentation, and then to justify the proposed solution to the user, to show that this particular product or service will be able to meet his needs.

Comparative study on personalized open-domain generative dialogue systems

Developing personalized dialogue systems is a monumental challenge in the rapidly evolving domain of Natural Language Processing (NLP). These systems are engineered to facilitate more natural, intuitive, and user-friendly interactions between humans and machines. Their theoretical significance extends to real-world applications, making them an invaluable asset in today's digital age. Over recent years, this area has garnered immense attention from researchers and academics, becoming a hotbed for innovation and study. This paper aims to contribute to this burgeoning field by conducting an in-depth comparative analysis of the current leading personalized language models. Specifically, this paper will scrutinize models grounded in deep learning algorithms and those that employ reinforcement learning techniques. This research objective is to dissect each approach's unique advantages and limitations. By doing so, the paper hopes to identify actionable avenues for improvement and optimization. Furthermore, the paper will offer a forward-looking perspective, outlining potential advancements and innovations that could shape the future landscape of personalized dialogue systems in Natural Language Processing.

Winnie: Task-Oriented Dialog System with Structure-Aware Contrastive Learning and Enhanced Policy Planning

Pre-trained encoder-decoder models are widely applied in Task-Oriented Dialog (TOD) systems on the session level, mainly focusing on modeling the dialog semantic information. Dialogs imply structural information indicating the interaction among user utterances, belief states, database search results, system acts and responses, which is also crucial for TOD systems. In addition, for the system acts, additional pre-training and datasets are considered to improve their accuracies, undoubtedly introducing a burden. Therefore, a novel end-to-end TOD system named Winnie is proposed in this paper to improve the TOD performance. First, to make full use of the intrinsic structural information, supervised contrastive learning is adopted to narrow the gap in the representation space between text representations of the same category and enlarge the overall continuous representation margin between text representations of different categories in dialog context. Then, a system act classification task is introduced for policy optimization during fine-tuning. Empirical results show that Winnie substantially improves the performance of the TOD system. By introducing the supervised contrastive and system act classification losses, Winnie achieves state-of-the-art results on benchmark datasets, including MultiWOZ2.2, In-Car, and Camrest676. Their end-to-end combined scores are improved by 3.2, 1.9, and 1.1 points, respectively.

Structure-Aware Multimodal Sequential Learning for Visual Dialog

With the ability to collect vast amounts of image and natural language data from the web, there has been a remarkable advancement in Large-scale Language Models (LLMs). This progress has led to the emergence of chatbots and dialogue systems capable of fluent conversations with humans. As the variety of devices enabling interactions between humans and agents expands, and the performance of text-based dialogue systems improves, there has been recently proposed research on visual dialog. However, visual dialog requires understanding sequences of pairs consisting of images and sentences, making it challenging to gather sufficient data for training large-scale models from the web. In this paper, we propose a new multimodal learning method leveraging existing large-scale models designed for each modality, to enable model training for visual dialog with small visual dialog datasets. The key ideas of our approach are: 1) storing the history or context during the progression of visual dialog in the form of spatiotemporal graphs, and 2) introducing small modulation blocks between modality-specific models and the graphs to align the semantic spaces. For implementation, we introduce a novel structure-aware cross-attention method, which retrieves relevant image and text knowledge for utterance generation from the pretrained models. For experiments, we achieved a new state-of-the-art performance on three visual dialog datasets, including the most challenging one COMET.

Cooper: Coordinating Specialized Agents towards a Complex Dialogue Goal

In recent years, there has been a growing interest in exploring dialogues with more complex goals, such as negotiation, persuasion, and emotional support, which go beyond traditional service-focused dialogue systems. Apart from the requirement for much more sophisticated strategic reasoning and communication skills, a significant challenge of these tasks lies in the difficulty of objectively measuring the achievement of their goals in a quantifiable way, making it difficult for existing research to directly optimize the dialogue procedure towards them. In our work, we emphasize the multifaceted nature of complex dialogue goals and argue that it is more feasible to accomplish them by comprehensively considering and jointly promoting their different aspects. To this end, we propose a novel dialogue framework, Cooper, which coordinates multiple specialized agents, each dedicated to a specific dialogue goal aspect separately, to approach the complex objective. Through this divide-and-conquer manner, we make complex dialogue goals more approachable and elicit greater intelligence via the collaboration of individual agents. Experiments on persuasion and emotional support dialogues demonstrate the superiority of our method over a set of competitive baselines. Our codes are available at https://github.com/YiCheng98/Cooper.

From Retrieval to Generation: A Simple and Unified Generative Model for End-to-End Task-Oriented Dialogue

Retrieving appropriate records from the external knowledge base to generate informative responses is the core capability of end-to-end task-oriented dialogue systems (EToDs). Most of the existing methods additionally train the retrieval model or use the memory network to retrieve the knowledge base, which decouples the knowledge retrieval task from the response generation task, making it difficult to jointly optimize and failing to capture the internal relationship between the two tasks. In this paper, we propose a simple and unified generative model for task-oriented dialogue systems, which recasts the EToDs task as a single sequence generation task and uses maximum likelihood training to train the two tasks in a unified manner. To prevent the generation of non-existent records, we design the prefix trie to constrain the model generation, which ensures consistency between the generated records and the existing records in the knowledge base. Experimental results on three public benchmark datasets demonstrate that our method achieves robust performance on generating system responses and outperforms the baseline systems. To facilitate future research in this area, the code is available at https://github.com/dzy1011/Uni-ToD.

Response Enhanced Semi-supervised Dialogue Query Generation

Leveraging vast and continually updated knowledge from the Internet has been considered an important ability for a dialogue system. Therefore, the dialogue query generation task is proposed for generating search queries from dialogue histories, which will be submitted to a search engine for retrieving relevant websites on the Internet. In this regard, previous efforts were devoted to collecting conversations with annotated queries and training a query producer (QP) via standard supervised learning. However, these studies still face the challenges of data scarcity and domain adaptation. To address these issues, in this paper, we propose a semi-supervised learning framework -- SemiDQG, to improve model performance with unlabeled conversations. Based on the observation that the search query is typically related to the topic of dialogue response, we train a response-augmented query producer (RA) to provide rich and effective training signals for QP. We first apply a similarity-based query selection strategy to select high-quality RA-generated pseudo queries, which are used to construct pseudo instances for training QP and RA. Then, we adopt the REINFORCE algorithm to further enhance QP, with RA-provided rewards as fine-grained training signals. Experimental results and in-depth analysis of three benchmarks show the effectiveness of our framework in cross-domain and low-resource scenarios. Particularly, SemiDQG significantly surpasses ChatGPT and competitive baselines. Our code is available at \url{https://github.com/DeepLearnXMU/SemiDQG}.

A Comprehensive Analysis of the Effectiveness of Large Language Models as Automatic Dialogue Evaluators

Automatic evaluation is an integral aspect of dialogue system research. The traditional reference-based NLG metrics are generally found to be unsuitable for dialogue assessment. Consequently, recent studies have suggested various unique, reference-free neural metrics that better align with human evaluations. Notably among them, large language models (LLMs), particularly the instruction-tuned variants like ChatGPT, are shown to be promising substitutes for human judges. Yet, existing works on utilizing LLMs for automatic dialogue evaluation are limited in their scope in terms of the number of meta-evaluation datasets, mode of evaluation, coverage of LLMs, etc. Hence, it remains inconclusive how effective these LLMs are. To this end, we conduct a comprehensive study on the application of LLMs for automatic dialogue evaluation. Specifically, we analyze the multi-dimensional evaluation capability of 30 recently emerged LLMs at both turn and dialogue levels, using a comprehensive set of 12 meta-evaluation datasets. Additionally, we probe the robustness of the LLMs in handling various adversarial perturbations at both turn and dialogue levels. Finally, we explore how model-level and dimension-level ensembles impact the evaluation performance. All resources are available at https://github.com/e0397123/comp-analysis.

TA&AT: Enhancing Task-Oriented Dialog with Turn-Level Auxiliary Tasks and Action-Tree Based Scheduled Sampling

Task-oriented dialog systems have witnessed substantial progress due to conversational pre-training techniques. Yet, two significant challenges persist. First, most systems primarily utilize the latest turn's state label for the generator. This practice overlooks the comprehensive value of state labels in boosting the model's understanding for future generations. Second, an overreliance on generated policy often leads to error accumulation, resulting in suboptimal responses when adhering to incorrect actions. To combat these challenges, we propose turn-level multi-task objectives for the encoder. With the guidance of essential information from labeled intermediate states, we establish a more robust representation for both understanding and generation. For the decoder, we introduce an action tree-based scheduled sampling technique. Specifically, we model the hierarchical policy as trees and utilize the similarity between trees to sample negative policy based on scheduled sampling, hoping the model to generate invariant responses under perturbations. This method simulates potential pitfalls by sampling similar negative policy, bridging the gap between task-oriented dialog training and inference. Among methods without continual pre-training, our approach achieved state-of-the-art (SOTA) performance on the MultiWOZ dataset series and was also competitive with pre-trained SOTA methods.

CHATBOT WITH FACIAL RECOGNITION BY USING AI

Creating a concept for a facial recognition chatbot is the main objective of the project. The work of chatbots in a variety of fields, such as marketing, education, healthcare, entertainment, has grown significantly in recent years. A chatbot is a piece of software that performs text or text-to-speech internet chats in place of face-to-face contact. Chatbots and other dialog systems are utilized for a number of purposes, including information gathering, request routing, and customer assistance. Some chatbot solutions look for generic keywords and respond with frequently used phrases that are pulled from a related library, while others utilize sophisticated AI, natural language processing, and intricate word categorization techniques. Here, we have added additional features including facial and language recognition. A facial recognition system is a technological device that uses a database of faces and digital images or video frames to identify a person's face. This kind of equipment locates and quantifies face features from a picture, usually to confirm users' identities through ID verification services.

OSTOD: One-Step Task-Oriented Dialogue with activated state and retelling response

As of present, the progress of conversational AI research has been greatly propelled by large-scale pre-trained language models. In particular, task-oriented dialogue systems have gained widespread attention owing to their immense potential in helping individuals accomplish diverse objectives, such as booking hotels, making restaurant reservations, and purchasing train tickets. In the past, task-oriented dialogue systems were typically viewed as a multi-step process that included spoken language understanding, dialogue state tracking, dialogue policy learning, and natural language generation. More recently, large-scale pre-trained language models enables the development of end-to-end neural pipeline task-oriented dialogue systems, which combine multiple steps into a single model, allowing for joint optimization and preventing error propagation. However, in order to explicitly retrieve information from databases to ensure the interpretability of the system, almost all end-to-end neural pipeline methods inevitably require predicting dialogue state as an intermediate result specialized for the domain or task, which results in significant challenges for generalization. To solve the problem above, we propose One-Step Task-Oriented Dialogue (OSTOD) in this paper, which models task-oriented dialogue by synchronously generating activated states and retelling responses, where activated states refer to slot values that contribute to database access, and retelling responses are system responses that contain activated state information. Specifically, first, automatic methods are designed to build data containing activated states and retelling responses. Then, a joint generation model that synchronously predicts activated states and retelling responses in a single step is proposed for task-oriented dialogue modelling. Based on empirical results obtained from the MultiWOZ 2.0 and MultiWOZ 2.1 datasets, our OSTOD model demonstrates comparable performance to state-of-the-art baselines. Moreover, our model exhibits exceptional generalization capabilities in few-shot learning and domain transfer scenarios.

Simulation modeling of the system of managing the psychophysiological state of a person based on non-verbal extralinguistic characteristics

Today, when artificial intelligence systems are being created, the time of constant dia-logue between humans and computer, there is a great demand for intelligent control systems and human-machine systems. These systems are very complex and require analysis and re-search. The publication is devoted to the simulation computer modeling of control of the psy-chological and physical state of a person. The basis of the human-computer dialogue system is the speech-to-text conversion sys-tem. The use of this system is the basis for the creation of computer language control systems. Unlike systems that work on the principle of identifying a single user response to a request and a proposal or a question from a password database, it is necessary to expand the lan-guage interface of the computer. It is necessary for the computer to accept commands from the operator and execute them only if the voice and its timbre match those registered in the database. Such a system will allow you to differentiate access to the computer and ensure se-curity for the user, preserve his privacy or save his personal data. After analyzing the speech activity of the operator, it became clear that it was necessary to develop a system of command control and «speech-to-text» conversion in the fused speech stream. This task is solved by comparing the words of the commands and is part of the problem of automatic recognition and understanding of the operator's human language. On the basis of such systems, a computer language control system can be created. The analysis has shown that unlike text-dependent identification systems, systems with a dialog method implement not only a one-time operator response to a query or question from the password database, but also the extension of its full-fledged language interface. The computer receives commands from the operator and executes them only if the voice matches the one registered in the database. This analysis is effective in the context of real-life speech activity of the operator. The literature review showed that for the development of sys-tems of command control and "speech-to-text" transformation in a stream of fused speech, it is necessary to use "mechanisms" of automatic recognition and understanding of the opera-tor's language.

A simple and efficient dialogue generation model incorporating commonsense knowledge

The performance of dialogue systems, artificial intelligence systems that communicate between a user and a machine, have rapidly improved owing to the development of the pre-trained language model that can perform well in various contexts. However, dialogue systems are often not well received by users because they generate universal and formulaic answers to user queries. As a result, users turn away from dialogue systems by reducing their interest in dialogue systems and lowering their expectations. To address this limitation, we propose a simple and efficient dialogue system that uses commonsense-based language models to facilitate more natural communication. Our model determines the context of human–machine conversations and then applies the relevant commonsense embedding to user queries. We quantitatively confirmed that our dialogue system performs significantly better on Korean datasets than non-commonsense-based dialogue systems.

KMc-ToD: Structure knowledge enhanced multi-copy network for task-oriented dialogue system

Task-oriented dialogue (ToD) system aims to assist users in completing various tasks, which has attracted great interest from researchers. However, the current models introduce delexicalization prepossessing to improve the generalization ability of the models, which makes it difficult to integrate dialogue slots into response generation, resulting in unsatisfactory performance for users. In this paper, we propose a structure knowledge-enhanced multi-copy network (KMc-ToD) for the ToD system, which uses the multi-copy mechanism to selectively copy dialogue slots from the dialogue history and schema graph into the response. Furthermore, to select the appropriate slots, we design the schema graph that helps the model understand the relation between different slots. Specifically, this graph adjusts the weights between nodes according to different utterances to prevent introducing noise. Experimental results on the MultiWOZ datasets show that our model achieves promising performance.

An open intent detection model optimized for datasets based on the Bert large model

Within current task-oriented dialogue systems, the focus of intent detection predominantly centers on closed domains. Nevertheless, in real-world usage scenarios, a substantial proportion of interactions fall into the open-domain category. User intentions frequently transcend predefined boundaries, giving rise to a multitude of out-of-domain intents, which pose a formidable challenge to existing models, ultimately leading to diminished recognition rates and accuracy. The demand for open intent detection models is increasing in today's society to address this issue effectively. This paper proposes a method to optimize datasets, thereby enhancing the training accuracy of open intent detection models. Specifically, this paper employs the Adaptive Decision Boundary Learning algorithm, which is currently popular in open intent detection. Leveraging this algorithm, this paper suggests using the K-means clustering algorithm to refine the intent labels within the dataset. This process helps identify and remove outliers in the dataset, making the distinction between known domain and open-domain intent labels more precise. Experimental results on two datasets, banking77 and stackoverflow, demonstrate the effectiveness of our approach in significantly improving model accuracy.

Promoting Unified Generative Framework with Descriptive Prompts for Joint Multi-Intent Detection and Slot Filling

Natural language understanding is a crucial aspect of task-oriented dialogue systems, encompassing intent detection (ID) and slot filling (SF). Conventional approaches for ID and SF solve the problems in a separate manners, while recent studies are now leaning toward joint modeling to tackle multi-intent detection and SF. Although the advancements in prompt learning offer a unified framework for ID and SF, current prompt-based methods fail to fully exploit the semantics of intent and slot labels. Additionally, the potential of using prompt learning to model the correlation between ID and SF in multi-intent scenarios remains unexplored. To address the issue, we propose a text-generative framework that unifies ID and SF. The prompt templates are constructed with label semantical descriptions. Moreover, we introduce an auxiliary task to explicitly capture the correlation between ID and SF. The experimental results on two benchmark datasets show that our method achieves an overall accuracy improvement of 0.4–1.5% in a full-data scenario and 1.4–2.7% in a few-shot setting compared with a prior method, establishing it as a new state-of-the-art approach.

The Latest Progress in Human-Computer Dialogue System

The Human-Computer dialogue system allows a computer to make conversations with humans through natural languages, and some designs can accomplish tasks given by humans. The development of the Human-Computer dialogue system can significantly expand the range of users for computers. The profound ELIZA Program designed scripts containing keywords and their corresponding rule-based sentence transformation to respond to the user. As the computation power increases, Human-Computer dialogue systems are handling more varied and complicated scenarios by introducing deeper and more complex artificial neural networks. The designer of a task-oriented system can choose between the pipeline method and the end-to-end method. The pipeline method is basically a pipeline that lets the user input go through three parts handling Natural Language Understanding, Dialogue Management, and Natural Language Generation, to generate an appropriate response back to the user. The end-to-end method, however, uses joint models to allow the parts to interact with others and get more efficient in handling the information. With the development of ChatGPT, more general language models and more varied methodologies are becoming more prevalent in dialogue systems.

MedChatZH: A tuning LLM for traditional Chinese medicine consultations

Generative Large Language Models (LLMs) have achieved significant success in various natural language processing tasks, including Question-Answering (QA) and dialogue systems. However, most models are trained on English data and lack strong generalization in providing answers in Chinese. This limitation is especially evident in specialized domains like traditional Chinese medical QA, where performance suffers due to the absence of fine-tuning and high-quality datasets. To address this, we introduce MedChatZH, a dialogue model optimized for Chinese medical QA based on transformer decoder with LLaMA architecture. Continued pre-training on a curated corpus of Chinese medical books is followed by fine-tuning with a carefully selected medical instruction dataset, resulting in MedChatZH outperforming several Chinese dialogue baselines on a real-world medical dialogue dataset. Our model, code, and dataset are publicly available on GitHub (https://github.com/tyang816/MedChatZH) to encourage further research in traditional Chinese medicine and LLMs.