Development Of Dialogue Systems Research Articles

A Modular Framework for Domain-Specific Conversational Systems Powered by Never-Ending Learning

Complex and long interactions (e.g., a change of topic during a conversation) justify the use of dialog systems to develop task-oriented chatbots and intelligent virtual assistants. The development of dialog systems requires considerable effort and takes more time to deliver when compared to regular BotBuilder tools because of time-consuming tasks such as training machine learning models and low module reusability. We propose a framework for building scalable dialog systems for specific domains using the semi-automatic methods of corpus, ontology, and code development. By separating the dialog application logic from domain knowledge in the form of an ontology, we were able to create a dialog system for the banking domain in the Portuguese language and quickly change the domain of the conversation by changing the ontology. Moreover, by using the principles of never-ending learning, unsupported operations or unanswered questions create triggers for system knowledge demand that can be gathered from external sources and added to the ontology, augmenting the system’s ability to respond to more questions over time.

Asymmetric communication: cognitive models of humans toward an android robot.

In the development of dialogue systems for android robots, the goal is to achieve human-like communication. However, subtle differences between android robots and humans are noticeable, leading even human-like android robots to be perceived differently. Understanding how humans accept android robots and optimizing their behavior is crucial. Generally, human customers have various expectations and anxieties when interacting with a robotic salesclerk instead of a human. Asymmetric communication arises when android robots treat customers like humans while customers treat robots as machines. Focusing on human-robot interaction in a tourist guide scenario, In this paper, we propose an asymmetric communication strategy that does not use estimation technology for preference information, but instead performs changing the agent's character in order to pretend to tailor to the customer. In line with this, we prepared an experimental method to evaluate asymmetric communication strategies, using video clips to simulate dialogues. Participants completed questionnaires without prior knowledge of whether the salesclerk was human-like or robotic. The method allowed us to assess how participants treated the salesclerk and the effectiveness of the asymmetric communication strategy. Additionally, during our demonstration in a dialogue robot competition, 29 visitors had a positive impression of the android robot's asymmetric communication strategy and reported a high level of satisfaction with the dialogue.

Ontology-driven development of dialogue systems

A new technique and its software implementation are presented to create a deeply semantically structured ontology using plain natural language text as input, without regular structure or any previous tagging and markup. The new approach is primarily aimed at highly inflectional languages, and is implemented for Ukrainian. The automatically created ontologies (in OWL) could be easily converted to other graph databases formats, such as Neo4j, and were successfully evaluated as valid ontologies using Protégé, RDFlib and Neo4j environments. An integrated approach is proposed for the development of natural language dialogue systems driven by the ontologyrelated graph database using the Cypher language for the formal queries. The original phrases are subject to a special method of semantic analysis, which determines the type of formal query to the database. The essence of the analysis is that the text of the user’s phrase goes through a series of checks. Based on their results, a set of basic templates for the formal requests are determined, as well as additional constructions that are attached to the basic template. Some of the checks may also return the notion of substitution to certain specified positions of the formal query. Formal queries can return both contexts and lists of ontology concepts. In addition to concepts, queries can also return information about specific semantic predicates that connect them, which simplifies the synthesis of natural language responses. The synthesis of answers is based on special templates, the choice of which is directly related to the corresponding template of the formal query.

Understanding computational dialogue understanding.

In this paper, we first explain why human-like dialogue understanding is so difficult for artificial intelligence (AI). We discuss various methods for testing the understanding capabilities of dialogue systems. Our review of the development of dialogue systems over five decades focuses on the transition from closed-domain to open-domain systems and their extension to multi-modal, multi-party and multi-lingual dialogues. From being somewhat of a niche topic of AI research for the first 40 years, it has made newspaper headlines in recent years and is now being discussed by political leaders at events such as the World Economic Forum in Davos. We ask whether large language models are super-parrots or a milestone towards human-like dialogue understanding and how they relate to what we know about language processing in the human brain. Using ChatGPT as an example, we present some limitations of this approach to dialogue systems. Finally, we present some lessons learned from our 40 years of research in this field about system architecture principles: symmetric multi-modality, no presentation without representation and anticipation feedback loops. We conclude with a discussion of grand challenges such as satisfying conversational maxims and the European Language Equality Act through massive digital multi-linguality-perhaps enabled by interactive machine learning with human trainers. This article is part of a discussion meeting issue 'Cognitive artificial intelligence'.

Dialogue Relation Extraction with Document-Level Heterogeneous Graph Attention Networks

We propose a heterogeneous graph attention network to address the problem of dialogue relation extraction. Compared with several popular sequence-based and graph-based models, our method shows superior performance on the benchmark dataset DialogRE. The implementation of this work can be found at https://github.com/declare-lab/dialog-HGAT Dialogue relation extraction aims to detect the relation between pairs of entities mentioned in a multi-party dialogue. It plays an essential role in understanding the deep logic of dialogues and facilitating the development of intelligent dialogue systems. We introduce a heterogeneous graph attention network to model the cross-sentence relations in a conversation. This heterogeneous graph attention network has modeled multi-type features of the conversation, such as utterance, word, speaker, argument, and entity type information. We compare our method with several popular baselines such as convolutional neural networks and long short-term memory, experimental results show our model outperforms the state-of-the-art method by 9.4%/7.8% F1 scores, and 6.6%/3.9% $$F1_c$$ scores in both validation and test sets with only 4.0M parameters. In this work, we present an attention-based heterogeneous graph network to deal with the dialogue relation extraction task in an inductive manner. Experimental results on the dataset DialogRE confirm the effectiveness of our method.

Turkish Data-to-Text Generation Using Sequence-to-Sequence Neural Networks

End-to-end data-driven approaches lead to rapid development of language generation and dialogue systems. Despite the need for large amounts of well-organized data, these approaches jointly learn multiple components of the traditional generation pipeline without requiring costly human intervention. End-to-end approaches also enable the use of loosely aligned parallel datasets in system development by relaxing the degree of semantic correspondences between training data representations and text spans. However, their potential in Turkish language generation has not yet been fully exploited. In this work, we apply sequence-to-sequence (Seq2Seq) neural models to Turkish data-to-text generation where the input data given in the form of a meaning representation is verbalized. We explore encoder-decoder architectures with attention mechanism in unidirectional, bidirectional, and stacked recurrent neural network (RNN) models. Our models generate one-sentence biographies and dining venue descriptions using a crowdsourced dataset where all field value pairs that appear in meaning representations are fully captured in reference sentences. To support this work, we also explore the performances of our models on a more challenging dataset, where the content of a meaning representation is too large to fit into a single sentence, and hence content selection and surface realization need to be learned jointly. This dataset is retrieved by coupling introductory sentences of person-related Turkish Wikipedia articles with their contained infobox tables. Our empirical experiments on both datasets demonstrate that Seq2Seq models are capable of generating coherent and fluent biographies and venue descriptions from field value pairs. We argue that the wealth of knowledge residing in our datasets and the insights obtained from this study hold the potential to give rise to the development of new end-to-end generation approaches for Turkish and other morphologically rich languages.

A dialogue system based on ontology automatically built through a natural language text analysis

An integrated approach is created to the development of natural-language dialogue systems driven by an ontological graph database. Ontology here has a defined regular structure that contains typed semantic relationships between concepts, as well as related contexts, which may also have a multilevel structure and additional typing. The ontology is created automatically due to the semantic analysis of a natural language using a specially developed original software, which is set up to work with inflected languages, in particular Ukrainian. The ontology description is serialized in OWL format. To work as part of the dialog system, the ontology is transferred to the graph database Neo4j. The Cypher language is used for formal queries. The original phrases of the user are subject to a special method of semantic analysis, which determines the type of formal query to the database. The essence of the analysis is that the text of the user phrase goes through a series of checks. Based on their results, a set of basic templates for formal requests is determined, as well as additional constructions that are attached to the basic template. Some of the checks may also return the notion of substitution to certain specified positions of the formal query. Formal queries can return both contexts and lists of ontology concepts. In addition to concepts, queries can also return information about specific semantic predicates that connect them, which simplifies the synthesis of natural language responses. The synthesis of answers is based on special templates, the choice of which is directly related to the corresponding template of the formal query.

A code-mixed task-oriented dialog dataset for medical domain

In the healthcare domain, medical and patient interactions form a crucial part of the diagnosis. Initially, the AI models developed for healthcare centered only on monolingual data. However, such models do not cater to the multilingual regions, where most conversations are Code-Mixed. We present the Code-Mixed Medical Task-Oriented Dialog Dataset to facilitate the research and development of Code-Mixed medical dialog systems. We analyzed the dataset using medical, conversational, and linguistic theories. The dataset contains 3005 Telugu–English Code-Mixed dialogs between patients and doctors with 29 k utterances covering ten specializations with an average code-mixing index (CMI) of 33.3%. We manually annotated the conversational dataset with intents and slot labels. We also present baselines to establish benchmarks on the dataset using existing state-of-the-art Natural Language Understanding (NLU) models. We improved the existing baselines using contextual ground truth intent labels and processing the slots as chunks. The data is made publically available.11https://github.com/suman101112/Code-Mixed-TOD-Medical-Dataset.git.

Do You Ever Get Off Track in a Conversation? The Conversational System’s Anatomy and Evaluation Metrics

Conversational systems are now applicable to almost every business domain. Evaluation is an important step in the creation of dialog systems so that they may be readily tested and prototyped. There is no universally agreed upon metric for evaluating all dialog systems. Human evaluation, which is not computerized, is now the most effective and complete evaluation approach. Data gathering and analysis are evaluation activities that need human intervention. In this work, we address the many types of dialog systems and the assessment methods that may be used with them. The benefits and drawbacks of each sort of evaluation approach are also explored, which could better help us understand the expectations associated with developing an automated evaluation system. The objective of this study is to investigate conversational agents, their design approaches and evaluation metrics. This approach can help us to better understand the overall process of dialog system development, and future possibilities to enhance user experience. Because human assessment is costly and time consuming, we emphasize the need of having a generally recognized and automated evaluation model for conversational systems, which may significantly minimize the amount of time required for analysis.

Development of Dialogue Management System for Banking Services

Rapid increase in conversational AI and user chat data lead to intensive development of dialogue management systems (DMS) for various industries. Yet, for low-resource languages, such as Azerbaijani, very little research has been conducted. The main purpose of this work is to experiment with various DMS pipeline set-ups to decide on the most appropriate natural language understanding and dialogue manager settings. In our project, we designed and evaluated different DMS pipelines with respect to the conversational text data obtained from one of the leading retail banks in Azerbaijan. In the work, the main two components of DMS—Natural language Understanding (NLU) and Dialogue Manager—have been investigated. In the first step of NLU, we utilized a language identification (LI) component for language detection. We investigated both built-in LI methods such as fastText and custom machine learning (ML) models trained on the domain-based dataset. The second step of the work was a comparison of the classic ML classifiers (logistic regression, neural networks, and SVM) and Dual Intent and Entity Transformer (DIET) architecture for user intention detection. In these experiments we used different combinations of feature extractors such as CountVectorizer, Term Frequency-Inverse Document Frequency (TF-IDF) Vectorizer, and word embeddings for both word and character n-gram based tokens. To extract important information from the text messages, Named Entity Extraction (NER) component was added to the pipeline. The best NER model was chosen among conditional random fields (CRF) tagger, deep neural networks (DNN), models and build in entity extraction component inside DIET architecture. Obtained entity tags fed to the Dialogue Management module as features. All NLU set-ups were followed by the Dialogue Management module that contains a Rule-based Policy to handle FAQs and chitchats as well as a Transformer Embedding Dialogue (TED) Policy to handle more complex and unexpected dialogue inputs. As a result, we suggest a DMS pipeline for a financial assistant, which is capable of identifying intentions, named entities, and a language of text followed by policies that allow generating a proper response (based on the designed dialogues) and suggesting the best next action.

The Effects of Physically Embodied Multiple Conversation Robots on the Elderly.

In recent years, communication robots aiming to offer mental support to the elderly have attracted increasing attention. Dialogue systems consisting of two robots could provide the elderly with opportunities to hold longer conversations in care homes. In this study, we conducted an experiment to compare two types of scenario-based dialogue systems with different types of bodies—physical and virtual robots—to investigate the effects of embodying such dialogue systems. Forty elderly people aged from 65 to 84 interacted with either an embodied desktop-sized humanoid robot or computer graphic agent displayed on a monitor. The elderly participants were divided into groups depending on the success of the interactions. The results revealed that (i) in the group where the robots responded more successfully with the expected conversation flow, the elderly are more engaged in the conversation with the physical robots than the virtual robots, and (ii) the elderly in the group in which robots responded successfully are more engaged in the conversation with the physical robots than those in the group in which the robots responded with ambiguous responses owing to unexpected utterances from the elderly. These results suggest that having a physical body is advantageous in promoting high engagement, and the potential advantage appears depending on whether the system can handle the conversation flow. These findings provide new insight into the development of dialogue systems assisting elderly in maintaining a better mental health.

HDRS: Hindi Dialogue Restaurant Search Corpus for Dialogue State Tracking in Task-Oriented Environment

Due to the rapid increase in the development of Task-oriented dialogue systems, the need for labelled dialogue corpus has become inevitable. For the Hindi language, there is no such dialogue corpus yet available. As a first attempt, we release a Hindi Dialogue Restaurant Search (HDRS) corpus and compare various state-of-the-art dialogue state tracking (DST) models on it. The corpus consists of 1.4 k human-to-human typed dialogues collected using Wizard-of-Oz paradigm. The paper starts with a brief description of the corpus by providing the details of features, corpus collection process and statistical analysis, then the performance of baseline NLU and DST models are investigated. Further, we experimented two categories of state-of-the-art belief state trackers: (1) Non-contextual pre-trained word embedding based DST models; (2) Contextual pre-trained BERT based DST models. All belief trackers follow a three-layered generic architecture. The category-1 models use the static domain ontology, while category-2 models have the capability to handle the dynamic ontology. The DST models are compared on joint-goal and turn-request accuracy. Global encoder and Slot-ATtentive decoders (GSAT) outperforms all the models with 83.25% joint-goal accuracy, followed by SUMBT.

Development of Natural Language Dialogue Software Systems

This paper discusses problems of natural language dialogue systems development. Here are considered different types of such program systems and their peculiarities based on our development experience. The systems discussed here are aimed at using graph ontological databases. Five basic types of natural language systems are highlighted: reference system, reference system with clarifying and technical questions, questionnaire system, active initiative system, control system. The first four of them are considered in detail. The reference system type, which is the basic type, is shown and discussed. Problems considered here are: natural language phrase to a formal query transformation, semantic analysis and answer formation. The ways are shown for these processes program implementation. For all the types the problem of the current dialogue subject keeping is considered, especially for an active initiative system where it is the most crucial. The different approaches are shown and compared for the principles of graph database (ontology) construction, which could be used in the dialogue systems.

Spoken dialogue system development at the dialogue robot competition

Spoken dialogue system development at the dialogue robot competition

Conversational AI: Dialogue Systems, Conversational Agents, and Chatbots

This book provides a comprehensive introduction to Conversational AI. While the idea of interacting with a computer using voice or text goes back a long way, it is only in recent years that this idea has become a reality with the emergence of digital personal assistants, smart speakers, and chatbots. Advances in AI, particularly in deep learning, along with the availability of massive computing power and vast amounts of data, have led to a new generation of dialogue systems and conversational interfaces. Current research in Conversational AI focuses mainly on the application of machine learning and statistical data-driven approaches to the development of dialogue systems. However, it is important to be aware of previous achievements in dialogue technology and to consider to what extent they might be relevant to current research and development. Three main approaches to the development of dialogue systems are reviewed: rule-based systems that are handcrafted using best practice guidelines; statistical data-driven systems based on machine learning; and neural dialogue systems based on end-to-end learning. Evaluating the performance and usability of dialogue systems has become an important topic in its own right, and a variety of evaluation metrics and frameworks are described. Finally, a number of challenges for future research are considered, including: multimodality in dialogue systems, visual dialogue; data efficient dialogue model learning; using knowledge graphs; discourse and dialogue phenomena; hybrid approaches to dialogue systems development; dialogue with social robots and in the Internet of Things; and social and ethical issues.

Методи та компоненти обробки природної мови

A dramatic change in the abilities of language models to provide state of the artaccuracy in a number of Natural Language Processing tasks is currently observed. These improvements open a lot of possibilities in solving NLP downstream tasks. Such tasks include machine translation, speech recognition, information retrieval, sentiment analysis, summarization, question answering, multilingual dialogue systems development and many more. Language models are one of the most important components in solving each of the mentioned tasks. This paper is devoted to research and analysis of the most adopted techniques and designs for building and training language models that show a state of the art results. Techniques and components applied in creation of language models and its parts are observed in this paper, paying attention to neural networks, embedding mechanisms, bidirectionality, encoder anddecoder architecture, attention and self-attention, as well as parallelization through using Transformer. Results: the most promising techniques imply pre-training and fine-tuning of a language model, attention-based neural network as a part of model design, and a complex ensemble of multidimensional embeddings to build deep context understanding. The latest offered architectures based on these approaches require a lot of computational power for training language model and it is a direction of further improvement.Ref. 49, pic. 13

Evaluating different Natural Language Understanding services in a real business case for the Italian language

Abstract In the last decade, the major private IT companies have developed lots of cloud platforms for Natural Language Understanding (NLU), that are widely used for research and commercial purposes. One of the main reasons for the success of NLU platforms is because they allow to simplify the Chatbots or Spoken Dialogue Systems (SDS) development and, in many cases, without knowledge about programming languages. In this paper, we present a general description and a taxonomy that brings together features and constraints of different cloud-based NLU services available on the market. Furthermore, we provide an evaluation and a comparison concerning the ability to recognise the underlying intents of different sentences. The sentences used are a collection of requests made by users in Italian on an e-learning platform. This analysis wants to help the company, owner of the e-learning platform, to understand in which NLU platform it is better to build a chatbot for the Italian language aiming at answering the most frequently asked questions.

Speech Technology Progress Based on New Machine Learning Paradigm.

Speech technologies have been developed for decades as a typical signal processing area, while the last decade has brought a huge progress based on new machine learning paradigms. Owing not only to their intrinsic complexity but also to their relation with cognitive sciences, speech technologies are now viewed as a prime example of interdisciplinary knowledge area. This review article on speech signal analysis and processing, corresponding machine learning algorithms, and applied computational intelligence aims to give an insight into several fields, covering speech production and auditory perception, cognitive aspects of speech communication and language understanding, both speech recognition and text-to-speech synthesis in more details, and consequently the main directions in development of spoken dialogue systems. Additionally, the article discusses the concepts and recent advances in speech signal compression, coding, and transmission, including cognitive speech coding. To conclude, the main intention of this article is to highlight recent achievements and challenges based on new machine learning paradigms that, over the last decade, had an immense impact in the field of speech signal processing.

An Efficient Framework for Development of Task-Oriented Dialog Systems in a Smart Home Environment.

In recent times, with the increasing interest in conversational agents for smart homes, task-oriented dialog systems are being actively researched. However, most of these studies are focused on the individual modules of such a system, and there is an evident lack of research on a dialog framework that can integrate and manage the entire dialog system. Therefore, in this study, we propose a framework that enables the user to effectively develop an intelligent dialog system. The proposed framework ontologically expresses the knowledge required for the task-oriented dialog system’s process and can build a dialog system by editing the dialog knowledge. In addition, the framework provides a module router that can indirectly run externally developed modules. Further, it enables a more intelligent conversation by providing a hierarchical argument structure (HAS) to manage the various argument representations included in natural language sentences. To verify the practicality of the framework, an experiment was conducted in which developers without any previous experience in developing a dialog system developed task-oriented dialog systems using the proposed framework. The experimental results show that even beginner dialog system developers can develop a high-level task-oriented dialog system.

A corpus of read and conversational Austrian German

This paper presents GRASS (Graz corpus of Read and Spontaneous Speech), the first large scale speech database for Austrian German with both read and conversational speech. In total, the corpus contains approximately 1900 min of speech in which 38 speakers produced more than 220,000 word tokens from 14,593 different word types. The corpus consists of three components. First, the Conversational Speech Component contains free conversations of one hour length between friends, colleagues, couples, or family members. Second, the Commands Component contains commands and keywords which were either read or elicited by pictures. Third, the Read Speech Component contains phonetically balanced sentences and digits. The speech of all components has been recorded at fullband quality in a soundproof recording-studio with head-mounted microphones, large-diaphragm microphones, a laryngograph, and with a video camera. The corpus was fully annotated at the orthographic level, and partly also at the segmental, sub-segmental and prosodic level. Our analysis of conversational speech characteristics such as overlapping speech, laughter, repetitions, hesitations and the use of colloquial and dialectal words allows us to conclude that the conversational speech material is highly casual in nature. The collected corpus provides conversational material for phoneticians and linguists interested in topics specific for Austrian German (e.g., pronunciation variability, prosody, syntax of spoken Austrian German), and for those studying talk in interaction in general (turn-taking, grounding, entrainment, extra-linguistic factors etc.). Furthermore, it is a valuable resource for speech technologists interested in the development of ASR and dialogue systems for different speaking styles of Austrian German.