Fluent Conversation Research Articles

Conceptualizing generative AI as style engines: Application archetypes and implications

The rise of generative AI has brought with it a surprising paradox: systems that excel at tasks once thought to be uniquely human, like fluent conversation or persuasive writing, while simultaneously failing to meet traditional expectations of computing, in terms of reliability, accuracy, and veracity (e.g., given the various issues with so-called ‘hallucinations’). We argue that, when generative AI is seen through a traditional computing lens, its development focuses on optimizing for traditional computing traits that remain in principle unattainable. This risks backgrounding what is most novel and defining about it. As probabilistic technologies, generative AIs do not store, in any traditional sense, any data or content. Rather, essential features of training data become encoded in deep neural networks as patterns, that become practically available as styles. We discuss what happens when the distinction between objects and their appearance dissolves and all aspects of images or text become understood as styles, accessible for exploration and creative combination and generation. For example, defining visual qualities of entities like ‘chair’ or ‘cat’ become available as ‘chair-ness’ or ‘cat-ness’ for creative image generation. We argue that, when understood as style engines, unique generative AI capabilities become conceptualized as complementing traditional computing ones. This will aid both computing practitioners and information systems researchers in reconciling and integrating generative AI into the traditional IS landscape. Our conceptualization leads us to propose four archetypes of generative AI application and use, and to highlight future avenues for information systems research made visible by this conceptualization, as well as implications for practice and policymaking.

Investigating the Learners’ Communicative Skills in ESL Classes

This article focuses on the communicative skills of learners in ESL classes. English conversation plays a crucial role in understanding each other and dealing with different kinds of uncovered problems. This means that not only do the soldiers that I teach need English for communication abroad, but Czech teachers also need English to communicate with them. It is always valuable to have a chat with them and ask questions to make sure our judgement colleagues are native speakers. At my work, I have an opportunity to encounter people from various parts of the world, and I am very interested in the way they use English as their mother tongue. These co-workers have a greater supporting role in our teaching process and were right. Even if they come from America, England, or Australia, and their accents differ, it is just a question of time for teachers and students to adjust to their speech and distinguish differences in pronunciation. Being able to keep a fluent conversation with a native speaker is viewed as the main goal of my students, which underlines the importance of speaking skills from a student´s point of view. Therefore, in my thesis, I decided to concentrate on communicative activities that might be helpful for English teachers and enhance their students´ communicative skills. My thesis is divided into two parts – theoretical and practical. In the theoretical part, I deal with the difference between speaking and conversation and explain its practical use in everyday life. I also aim to highlight the importance of motivation in teaching practice and the techniques and approaches used to enhance students´ desire to speak a foreign language. While dealing with communicative activities, I focused on distinguishing them either as accuracy or fluency activities and provided definitions of these terms. The role of a teacher is also discussed, and useful ideas are suggested to make the speaking lesson as effective as possible. The practical part consists of a set of pre-taught communicative activities described minutely and their evaluation for other teachers who would like to apply them in their speaking lessons.

Expressive Sign Language Recognition using CNN

A major shortcoming in our society is a social barrier between the differently abled members of the society and the abled folks. One of the most important aspects of human beings, being regarded as social animals, is in fact communication. Communication is also a major obstacle faced by the hearing and vocal disabilities people. This inability to communicate leads to frequent problems and hinders the daily activities of a person with hearing and vocal disabilities. The underlying reason for this disparity is that abled folks don’t learn and aren’t taught Sign Language which is the main means of communication for a person with hearing and vocal disabilities. Thus, abled folks are incapable of having a normally fluent conversation with these different sections of the society. Consequently, in a verbal exchange among hearing and speech impaired individuals and an able person the convenience of communique and consequently the consolation degree is hampered. So, in our project, we have proposed a cost-efficient solution to overcome this communication barrier. This solution can be easily used by everyone and can also be, with some modifications, made to work on most platforms which have a camera module. Our approach uses the integrated camera module to capture real time hand gestures based on hand key points or landmarks and the algorithm using machine learning techniques, displays the alphabet that the gesture is representing. KEY WORDS: Convolution Neural Network, Transfer Learning, Deep Learning, Artificial Neural Network, Support Vector Machines

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.

Recommendations for promoting user agency in the design of speech neuroprostheses.

Brain-computer interfaces (BCI) that directly decode speech from brain activity aim to restore communication in people with paralysis who cannot speak. Despite recent advances, neural inference of speech remains imperfect, limiting the ability for speech BCIs to enable experiences such as fluent conversation that promote agency - that is, the ability for users to author and transmit messages enacting their intentions. Here, we make recommendations for promoting agency based on existing and emerging strategies in neural engineering. The focus is on achieving fast, accurate, and reliable performance while ensuring volitional control over when a decoder is engaged, what exactly is decoded, and how messages are expressed. Additionally, alongside neuroscientific progress within controlled experimental settings, we argue that a parallel line of research must consider how to translate experimental successes into real-world environments. While such research will ultimately require input from prospective users, here we identify and describe design choices inspired by human-factors work conducted in existing fields of assistive technology, which address practical issues likely to emerge in future real-world speech BCI applications.

The Effects of Language Interference in Learning English Syntax for Spanish Native Speakers: A Case Study on a Mexican Context

Language interference is a relevant topic for second language acquisition because most language learners face this phenomenon when learning a new language. This study followed a social constructivist philosophy, focusing on indictive and qualitative methods. This case study was conducted in Leon, Guanajuato, Mexico. In this case study, I explored the experiences of two young-adults language learners from the center of Guanajuato, Mexico. These participants have been taking formal English lessons for more than twelve years. However, these learners are still not able to have fluent conversations in English. The instruments used for this research were questionnaires, interviews, and recorded conversations. After conducting a thematic analysis, the result indicates that the main issue these participants are facing is the use of translation as a strategy and the concern the participants have of committing mistakes.

The Social Representations of Zoo Goers toward Crocodiles and Turtles: Structural Analysis and Implications for Conservation

Zoos have changed dramatically over the last century and today attract millions of people worldwide, being places where visitors can closely watch wildlife and learn about the species on display. Although present at most zoos, reptiles are challenging animals in terms of visitor interest and engagement, as some species do not fit aesthetic standards from the human standpoint, have culturally negative perceptions or generate aversive emotions. By studying zoo visitors’ social representations of crocodiles and turtles, we aimed to detail their structures, as well as identifying their prototypical elements that help to understand their emotional and cognitive framing. The findings show the crocodile’s prototypical image as a big, fearsome predator with teeth as its main physical attribute. Male visitors showed a more emotional perception of this animal. The turtle’s prototypical image is a slow, hard-shelled ancestral sea animal with a neutral-to-positive set of traits, with no particular differences between genders. Our results shed a more detailed light on some of the social constructs that make up the mental images of these animals, which can help the zoological community direct communication toward a more fluent conversation between stakeholders toward conservation.

PERSONALIZED INTELLIGENT VOICE ASSISTANT ANNA FOR SPEECH RECOGNITION USING NATURAL LANGUAGE PROCESSING

As the use of artificial intelligence increases, this technology is more advanced than ever. Artificial intelligence is integrated into the components of life. What was unthinkable 10 years ago is now a casual act, with fluent conversations with computers. This was made possible by concepts of machine learning and speech recognition. That is, you can only run and complete user-designed quests. Plus, get started with audiobooks, SIM tracking, and Google Meet right away. Even if we combine voice technology with human-like conversations, There is no need to talk to our devices and have a conversation to complete our task of choice. The real focus should be on unspoken commands in completing the user’s task as quickly as possible. Voice UI is getting popular as it makes the work easy and when it is integrated with a device it will become a good user product. This is why we developed a personalized user-designated model that fulfills all the features and provides a good experience

Machine learning approaches for biomolecular, biophysical, and biomaterials research.

A fluent conversation with a virtual assistant, person-tailored news feeds, and deep-fake images created within seconds-all those things that have been unthinkable for a long time are now a part of our everyday lives. What these examples have in common is that they are realized by different means of machine learning (ML), a technology that has fundamentally changed many aspects of the modern world. The possibility to process enormous amount of data in multi-hierarchical, digital constructs has paved the way not only for creating intelligent systems but also for obtaining surprising new insight into many scientific problems. However, in the different areas of biosciences, which typically rely heavily on the collection of time-consuming experimental data, applying ML methods is a bit more challenging: Here, difficulties can arise from small datasets and the inherent, broad variability, and complexity associated with studying biological objects and phenomena. In this Review, we give an overview of commonly used ML algorithms (which are often referred to as "machines") and learning strategies as well as their applications in different bio-disciplines such as molecular biology, drug development, biophysics, and biomaterials science. We highlight how selected research questions from those fields were successfully translated into machine readable formats, discuss typical problems that can arise in this context, and provide an overview of how to resolve those encountered difficulties.

A systematic review and Bayesian meta-analysis of the development of turn taking in adult-child vocal interactions.

Fluent conversation requires temporal organization between conversational exchanges. By performing a systematic review and Bayesian multi-level meta-analysis, we map the trajectory of infants' turn-taking abilities over the course of early development (0 to 70months). We synthesize the evidence from 26studies (78 estimates from 429 unique infants, of which at least 152 are female) reporting response latencies in infant-adult dyadic interactions. The data were collected between 1975 and 2019, exclusively in North America and Europe. Infants took on average circa 1s to respond, and the evidence of changes in response over time was inconclusive. Infants' response latencies are related to those of their adult conversational partners: an increase of 1s in adult response latency (e.g., 400 to 1400ms) would be related to an increase of over 1s in infant response latency (from 600 to 1857ms). These results highlight the dynamic reciprocity involved in the temporal organization of turn-taking. Based on these results, we provide recommendations for future avenues of enquiry: studies should analyze how turn-by-turn exchanges develop on a longitudinal timescale, with rich assessment of infants' linguistic and social development.

Dementia with aphasia and mirror phenomenon: examination of the mechanism using neuroimaging and neuropsychological findings: a case report

BackgroundAphasia often appears in persons living with dementia; however, aphasia and the mirror phenomenon are rarely present at the same time.Case presentationHere, we report a case of fluent conversation with a person in a mirror or a magazine, and examine the underlying mechanism using brain imaging and neuropsychological findings. We found that the appearance of the mirror phenomenon may be associated with a visuospatial dysfunction caused by a decreased function of the posterior region of the right temporal and parietal lobe. Moreover, active talking to a person in a mirror or a person in a magazine could be associated with disinhibition caused by a decline in bilateral frontal lobe function.ConclusionsThis case represents a very valuable and interesting presentation because it is the first report of a long-term follow-up of the course of dementia using neurological imaging, and of the neuropsychological analysis of the mechanism of conversation with a mirror image combined with aphasia.

Perception of influences on the diet of elderly people from the Lan-C sector of Hualpén commune, Chile

Objective: To interpret the perception of influences on the diet of elderly people from the Lan-C sector of Hualpén commune, Chile, during the year 2016. Materials and methods: In order to accomplish the objective, researchers used a qualitative design with an interpretative phenomenological approach. The qualitative sampling was performed by the snowball method. The sample size was not determined a priori, but through theoretical saturation; that is, when interviews did not provide new data to the research, which happened in the fifth interview. Subjects included in the study fulfilled the screening criteria, such as being over 60 years old, not having cognitive damage that prevented a fluent conversation, not being diagnosed with any form of dementia, living in the Lan-C sector of Hualpén commune, and signing an informed consent of the research with which ethical aspects of the study were controlled. Data collection was performed through semi-structured interviews. The results were coded and analyzed through the content analysis method. Results: Interviewees mentioned that their diet was influenced by their family, emotions and economic situation. The family, especially their children, had a positive influence on their diet by selecting the food they were to consume, preparing their meals, and maintaining or improving their health through the diet. They also stated that the economic situation affected their diet and, although they had a limited income, they favored the purchase of healthy food which would allow them to keep an adequate quality of life. Regarding their emotions, they especially refer to memories of their childhood and youth, their mother and special dates which have influence on their diet. Other factors, such as advertisement or friends, were not identified as conditioning their eating behavior. Conclusion: There are some factors that affect the diet of elderly people, such as their emotions, economic situation and family.

A conversation analytic study of patterns of overlapping talk in conversations between individuals with dementia and their frequent communication partners.

Difficulty participating in conversation is commonly experienced by individuals with dementia, secondary to cognitive and language deficits. Frequent communication partners (FCPs), however, report being largely unaware of how to support their conversation partners with dementia during conversation. In particular, taking a turn appropriately may be difficult for either partner due to trouble predicting a partner's behaviour and, hence, difficulty with timing conversational turns appropriately, potentially resulting in overlapping talk. To investigate the patterns of overlapping talk in the interaction between individuals with dementia and their FCPs. Three participants with moderate-severe dementia participated in conversation with an FCP. Ten minutes of 'casual' and 'task-oriented' conversation were audio- and video-recorded. Patterns of overlapping talk were investigated using conversation analytic methods. Overlapping talk was a consistent feature of all three dyadic interactions during both social and task-oriented talk. All participants exhibited competitive and non-competitive forms of overlapping talk. The data reveal that FCPs commonly yielded their own turns when overlapped by a partner in order to create opportunities for their partners with dementia to communicate. Participants with dementia demonstrated some retained pragmatic abilities, both using continuers and yielding the floor to their partner when competitively overlapped in order to encourage a speaker to continue. These findings contribute to the understanding of the impact of dementia on the maintenance of sensitivity to the sequential aspects of everyday talk. From a clinical perspective, these findings can inform the training of FCPs about retained abilities and evidence-based support strategies, equipping them with knowledge and skills to structure and maintain fluent conversation.

Conversational gaze mechanisms for humanlike robots

During conversations, speakers employ a number of verbal and nonverbal mechanisms to establish who participates in the conversation, when, and in what capacity. Gaze cues and mechanisms are particularly instrumental in establishing the participant roles of interlocutors, managing speaker turns, and signaling discourse structure. If humanlike robots are to have fluent conversations with people, they will need to use these gaze mechanisms effectively. The current work investigates people's use of key conversational gaze mechanisms, how they might be designed for and implemented in humanlike robots, and whether these signals effectively shape human-robot conversations. We focus particularly on whether humanlike gaze mechanisms might help robots signal different participant roles, manage turn-exchanges, and shape how interlocutors perceive the robot and the conversation. The evaluation of these mechanisms involved 36 trials of three-party human-robot conversations. In these trials, the robot used gaze mechanisms to signal to its conversational partners their roles either of two addressees, an addressee and a bystander, or an addressee and a nonparticipant. Results showed that participants conformed to these intended roles 97% of the time. Their conversational roles affected their rapport with the robot, feelings of groupness with their conversational partners, and attention to the task.

Disrupting the flow: How brief silences in group conversations affect social needs

We all know the awkward feeling when a conversation is disrupted by a brief silence. This paper studies why such moments can be unsettling. We suggest that silences are particularly disturbing if they disrupt the conversational flow. In two experiments we examined the effects of a single brief instance of silence on social needs, perceived consensus, emotions, and rejection. Study 1 demonstrated that fluent conversations are associated with feelings of belonging, self-esteem, and social validation. If a brief silence disrupts this fluency, negative emotions and feelings of rejection arise. Study 2 replicated these effects in a more realistic setting, and showed that the effects of a brief silence are considerable despite participants' unawareness of the silence. Together, results show that conversational flow induces a sense of belonging and positive self-esteem. Moreover, this research suggests an implicit route to social validation, in which consensus is inferred from fluent group conversation.

Once more with feeling: Utterance and social structure

AbstractIn this paper, I examine three theoretical sources on the relationship between language use and social structure. I identify common themes which link the concepts of symbolic capital (Bourdieu), social semiotic (Halliday), and ideologeme (Medvedev) and which lead to a critique of the structuralist tradition of semiotic theory. I explore the significance of speech prosody as an integral part of speech communication, and suggest that the utterance needs to be seen as a complex whole in which the structural and dynamic elements of speech are functionally combined. I illustrate this position through the analysis of a sequence of naturalistic dialogue, highlighting aspects of prosodic orientation between speakers and discussing its semiotic significance. I suggest that the mediating concept of speech genre can be used to understand how fluent conversation has both a structured and an improvisatory character. I conclude by suggesting that the activity of producing an utterance in everyday speech may be seen in terms of a musical analogy as performing a variation on a traditional theme. In closing, I identify a number of topics in this field on which our knowledge remains underdeveloped and indicate lines of enquiry for future research.

A theory of lexical access in speech production.

Preparing words in speech production is normally a fast and accurate process. We generate them two or three per second in fluent conversation; and overtly naming a clear picture of an object can easily be initiated within 600 msec after picture onset. The underlying process, however, is exceedingly complex. The theory reviewed in this target article analyzes this process as staged and feed-forward. After a first stage of conceptual preparation, word generation proceeds through lexical selection, morphological and phonological encoding, phonetic encoding, and articulation itself. In addition, the speaker exerts some degree of output control, by monitoring of self-produced internal and overt speech. The core of the theory, ranging from lexical selection to the initiation of phonetic encoding, is captured in a computational model, called WEAVER++. Both the theory and the computational model have been developed in interaction with reaction time experiments, particularly in picture naming or related word production paradigms, with the aim of accounting for the real-time processing in normal word production. A comprehensive review of theory, model, and experiments is presented. The model can handle some of the main observations in the domain of speech errors (the major empirical domain for most other theories of lexical access), and the theory opens new ways of approaching the cerebral organization of speech production by way of high-temporal-resolution imaging.

Some advantages of video conferencing over high-quality audio conferencing: fluency and awareness of attentional focus

There are many commercial systems capable of transmitting a video image of parties in a conversation over a digital network. Typically, these have been used to provide facial images of the participants. Experimental evidence for the advantages of such a capability has been hard to find. This paper describes two experiments that demonstrate significant advantages for video conferencing over audio-only conferencing, in the context of a negotiation task using electronically shared data. In the video condition there was a large, high-quality image of the head and upper torso of the participant(s) at the other end of the link and high-quality sound. For the audio-alone condition the sound was the same but there was no video image. The criteria by which these two communication conditions were compared were not the conventional measures of task outcome. Rather, measures relating to conversational fluency and interpersonal awareness were applied. In each of the two experiments, participants completed the same task with data presented by a shared editor. In Experiment 1, they worked in pairs and in Experiment 2 they worked at quartets with two people at each end of the link. Fluency was assessed from transcripts in terms of length of utterance, overlapping speech and explicit questions. Only the latter measure discriminated between the two communication conditions in both experiments. The other measures showed significant effects in Experiment 2 but not in Experiment 1. Given this pattern of results it is concluded that video can result in more fluent conversation, particularly where there are more than two discussants. However, in the case of dyadic conversation auditory cues to turn taking, etc., would seem to suffice. In both experiments there was a large and significant effect on interpersonal awareness as assessed by ratings of the illusion of presence, and most clearly, awareness of the attentional focus of the remote partner (s). In Experiment 2, the ratings for the remote partners were similar to those for the co-located discussants, demonstrating the effectiveness of the video link with regard to these subjective scales.

Targets and results from phase one and two of the fifth generation computer systems study

This article presents an introduction to the fifth generation computer systems. Fifth generation computers are under development to meet the knowledge information processing needs for the advanced information-oriented society of the 1990s. This clearly requires overcoming the technical limitations of conventional computers. A fifth generation computer requires problem solving and inference functions for inference of solutions to problems from accumulated knowledge, knowledge base management functions for acquirement and retrieval of knowledge, intelligent interface functions for fluent conversations using natural language, drawings, images, etc., and intelligent programming functions for automatic conversion of problem specifications submitted in natural language, drawing etc., into efficient programs. Fifth generation computers with these functions will have far-reaching impact finding applications in all areas of industry and society. However, research and development (R&D) of these computers requires an extremely wide variety of leading-edge technology, and the risks are high. Accordingly, a long-term R&D plan was established for this project that extends over ten years with the initial stage lasting from 1982 to 1984, the intermediate stage from 1985 to 1988, and the final stage will begin in 1989 and the project will end in 1991.