Conversational Context Research Articles

Know your enemy: Conversational agents for security, education, training, and awareness at scale

Social engineering attacks are widespread, costly, and challenging to mitigate through technical means. Companies allocate considerable resources toward security education, training, and awareness. However, the cost of these efforts is high, and the efficacy is questionable, particularly for defending against social engineering attacks. To improve security training for employees against social engineering at scale, we have designed a conversational agent to simulate a person becoming the victim of a social engineering attack. We build on Utility theory and previous social engineering, adversarial thinking, and conversational agent research. We designed a conversational agent to allow humans to learn and practice social engineering techniques in a fun and engaging setting. Users learn how an attacker might use these techniques by performing them. We present the design of a conversational agent for social engineering training and demonstrate a process to measure adversarial thinking in a computer-mediated conversational context. We showed that the conversational agent performs well and accurately measures adversarial thinking through two studies.

Embodying Ethiopia’s Global Golden Age on the Muslim-Christian Frontier: The Allure of Glass Beads

The period between AD 700 and 1500 has been recently labeled as “Africa’s global Golden Age.” This is particularly true for the Shay communities living on the Muslim-Christian frontier in the ninth to fourteenth century AD. Located in the center of the Ethiopian highlands, the Shay faced the expansion of the Christian kingdoms and the advance of the Muslim polities. In an increasingly violent context of religious conversion and war between the two religious powers, the Shay stressed their independence by burying their deceased in collective structures, contrary to the mortuary practices of both Christians and Muslims, and by including precious local and global grave goods in their tombs. The laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) analysis of 34 glass beads shows how the Shay communities benefited from the Islamic global trade routes at the time, particularly the Middle East, Egypt, and the Indo-Pacific networks. This article examines the crucial role of global glass beads in the construction of a trans-corporeal landscape among the Shay that served the emergence and consolidation of the social self as a collective identity against their Christian and Muslim neighbors.

Linguistic features and pragmatic functions of direct reported speech in Italian troubles telling sequences

This article aims to investigate possible linguistic features and pragmatic functions of Direct Reported Speech (DRS) in a specific conversational context, which seems to favour the emerging of this phenomenon: the trouble telling sequence. The study, conducted on a corpus of naturally occurring interactions in Italian, describes some characteristics of this linguistic device from both a qualitative and a quantitative point of view, by analysing how and for what purposes it can be used. DRS produced by the Troubles Teller mostly aims to account for their point of view, to complain and criticise, to express feelings and emotions and to anticipate future behaviour; the Trouble Recipient uses DRS mainly to express support towards the conversational partner and to give advice.

The Pragmatic Effects of the Second Language on Native Language’s Acquisition and Application in Conversational Context

With the development of frequent communication between countries, the present paper aims to explore the pragmatic effects of the second language on native language’s acquisition and application, especially in the conversational context. During the application of a foreign language, the Second Language (L2) is also exerting a latent influence on the use of the First Language (L1), which is named as "reverse transfer." The data is collected from a questionnaires and transcripts of recorded audios. There are 100 participants involved, who were seperated into three groups. After the comparison between groups, this research conducted a quantitative analysis of the data. Main findings were: 1) The influence of the L2 on the L1 can be understood as the application of the native language is influenced by the foreign language in the process of SLA. 2) The reverse transfer is not outstanding at the early stage of the beginning of language study. However, there will be more impact of L2 on L1 when the learner achieves proficiency at some level of their L2. 3) At present, the study of the backward effects of a second language on a native language has attracted extensive attention to language acquisition and application in the pragmatics field. However, scholars have held disparate views on the extent to which the second language affects the first language. Therefore, the study of the reverse effect of the second language on the native language can help bilingual learners better acquire and improve their bilingual ability. Future researchers are likely to be inspired to work on the application of the second language in manifold categories of pragmatics, which can have further preventive measures in infection control of bilingual speech-language pathologies, such as communication disorders and cognitive-communication disorders. Furthermore, it can better avoid communicative lingerings, cross-language intrusion, and first language attrition.

Social Media Network Analysis of Academic Urologists’ Interaction Within Twitter Microblogging Environment

ObjectiveTo characterize academic urology Twitter presence and interaction by subspecialty designation.MethodsUsing Twitter application programming interface of available data, 94000 specific tweets were extracted for the analysis through the Twitter Developer Program. Academic urologists were defined based on American Urological Association (AUA) residency program registration of 143 residency programs, with a total of 2377 faculty. Two of 3-factor verification (name, location, specialty) of faculty Twitter account was used. Additional faculty information including sex, program location, and subspecialty were manually recorded. All elements of microblogging were captured through Anaconda Navigator. Analyzed tweets were further evaluated using natural language processing for sentiment association, mentions, and quote tweeted and retweeted. Network analysis based on interactions of academic urologist within specialty for given topic were analyzed using D3 in JavaScript. Analysis was performed in Python and R.ResultsWe identified 143 residency programs with a total of 2377 faculty (1975 men and 402 women). Among all faculty, 945 (39.7%) had registered Twitter accounts, with the majority being men (759 [80.40%] versus 185 [19.60%]). Although there were more male academic urologists across programs, women within academic urology were more likely to have a registered Twitter account overall (46% versus 38.5%) compared with men. When assessing registered accounts by sex, there was a peak for male faculty in 2014 (10.05% of all accounts registered) and peak for female faculty in 2015 (2.65%). There was no notable change in faculty account registration during COVID-19 (2019–2020). In 2022, oncology represented the highest total number of registered Twitter users (225), with the highest number of total tweets (24622), followers (138541), and tweets per user per day (0.32). However, andrology (50%) and reconstruction (51.3%) were 2 of the highest proportionally represented subspecialties within academic urology. Within the context of conversation surrounding a specified topic (#aua21), female pelvic medicine and reconstructive surgery (FPMRS) and endourology demonstrated the total highest number of intersubspecialty conversations.ConclusionsThere is a steady increase in Twitter representation among academic urologists, largely unaffected by COVID-19. While urologic oncology represents the largest group, andrology and reconstructive urology represent the highest proportion of their respective subspecialties. Interaction analysis highlights the variant interaction among subspecialties based on topic, with strong direct ties between endourology, FPMRS, and oncology.

SPK-CG: Siamese Network based Posterior Knowledge Selection Model for Knowledge Driven Conversation Generation

Building a human-computer conversational system that can communicate with humans is a research hotspot in the field of artificial intelligence. Traditional dialogue systems tend to produce irrelevant and non-information responses, which reduce people’s interest in engaging in a conversation. This often leads to boring conversations. To alleviate this problem, many researchers use external knowledge to assist conversation generation. The accuracy of knowledge selection is the prerequisite to ensure the quality of knowledge conversation. This approach has worked positively to a certain extent, but generally only searches knowledge information based on entity words themselves, without considering the specific conversation context. Therefore, if irrelevant knowledge is retrieved, the quality of conversation generation will be reduced. Motivated by this, we propose a novel neural knowledge-based conversation generation model, named Siamese Network based Posterior Knowledge Selection Model for Knowledge Driven Conversation Generation (SPK-CG) . We have designed a novel knowledge selection mechanism to obtain knowledge information that is highly relevant to the context of the conversation. Specifically, the posterior knowledge distribution is used as a soft label to make the prior distribution consistent with the posterior distribution in the training process. At the same time, in order to narrow the gap between prior and posterior distributions and improve the accuracy of knowledge selection, we leverage siamese network and design multi-granularity matching module for knowledge selection. Compared with previous knowledge-based models, our method can select more appropriate knowledge and use the selected knowledge to generate responses that are more relevant to the conversation context. Extensive automatic and human evaluations demonstrate that our model has advantages over previous baselines.

Measuring “sustainable development” in vernacular settlements: a case study Behramkale, Türkiye

PurposeThe main objective of this paper is to critically assess sustainable development in the context of Behramkale, a vernacular village in Türkiye.Design/methodology/approachVernacular Heritage Sustainable Architecture analysis framework has been adopted to understand and assess vernacular architecture and sustainable development in Behramkale.FindingsThe vernacular design of the old Behramkale settlement has shown more sustainable characteristics as compared to the new development area. Key findings show that trade-offs were made with respect to environmental and sociocultural aspects of sustainable development to achieve economic sustainability.Research limitations/implicationsFuture research with more in-depth interviews would be helpful to find out the inhabitants’ response to the conservation practices.Practical implicationsBased on the research conducted, life cycle analysis and sustainable strategies of vernacular settlements can be useful tools to design, develop and improve old settlements, as well as newly established settlements.Social implicationsKey lessons learned from conservation practices can help to identify well-adapted solutions to respond to the needs of local communities in Türkiye and similar vernacular settlements in the Mediterranean region.Originality/valueThis paper critically assesses sustainable development in the context of vernacular architecture, heritage conservation and rural sustainability. Conservation practices in Türkiye are evaluated deeply as there is limited research in this field within the Mediterranean heritage conversation and sustainable development context.

Quantifying the interplay of conversational devices in building mutual understanding.

Humans readily engage in idle chat and heated discussions and negotiate tough joint decisions without ever having to think twice about how to keep the conversation grounded in mutual understanding. However, current attempts at identifying and assessing the conversational devices that make this possible are fragmented across disciplines and investigate single devices within single contexts. We present a comprehensive conceptual framework to investigate conversational devices, their relations, and how they adjust to contextual demands. In two corpus studies, we systematically test the role of three conversational devices: backchannels, repair, and linguistic entrainment. Contrasting affiliative and task-oriented conversations within participants, we find that conversational devices adaptively adjust to the increased need for precision in the latter: We show that low-precision devices such as backchannels are more frequent in affiliative conversations, whereas more costly but higher-precision mechanisms, such as specific repairs, are more frequent in task-oriented conversations. Further, task-oriented conversations involve higher complementarity of contributions in terms of the content and perspective: lower semantic entrainment and less frequent (but richer) lexical and syntactic entrainment. Finally, we show that the observed variations in the use of conversational devices are potentially adaptive: pairs of interlocutors that show stronger linguistic complementarity perform better across the two tasks. By combining motivated comparisons of several conversational contexts and theoretically informed computational analyses of empirical data the present work lays the foundations for a comprehensive conceptual framework for understanding the use of conversational devices in dialogue. (PsycInfo Database Record (c) 2022 APA, all rights reserved).

Executive functions are modulated by the context of dual language use: diglossic, bilingual and monolingual older adults

AbstractStudies investigating the role of dual language use in modulating executive functions have reported mixed results, with some studies reporting benefits in older adults. These studies typically focus on bilingual settings, while the role of dual language use in diglossic settings is rarely investigated. In diglossia, the two language varieties are separated by context, making it an ideal test case for the effects on cognition of Single Language Contexts, as defined by the Adaptive Control Hypothesis (Green & Abutalebi, 2013). We compare the performances of three groups of older adults, Arab diglossics (n = 28), bilinguals (n = 29), and monolinguals (n = 41), on the Flanker and Stroop tasks, measuring inhibition abilities, and the Color-shape task, measuring switching abilities. We report a diglossic benefit in inhibition as measured by the Flanker task only, and no benefits for the bilingual group. These findings are discussed with reference to conversational contexts in dual language use.

A DESCRIPTIVE STUDY ON STUDENTS’ SPEAKING ABILITY IN “DAILY CONVERSATION” CONTEXT AT HIGHER EDUCATION

Speaking is one of English skill that has been learnt from junior high school until higher education. It means that students in higher education should have good capability in doing speaking in all contexts, include in daily conversation context. For this reason, this study was designed to determine the extent to which students' ability in speaking especially in daily conversation context. This study took 22 students in Educational Faculty of Tompotika Luwuk University as the participants. Descriptive qualitative is a method that chosen in this study. The data have been collected from observation and documentation. Observation and documentation collected when the students doing conversation. The result of this study shows that the students’ speaking ability in daily conversation context can be categorized into “poor”. It was proven by the students who got only 18,18% in good category, 22,72% in average category, 40,92% in poor category, and 18,2% in very poor category. Finally, it can be conclude that the students’ speaking ability in daily conversation context at higher education of Educational Faculty of Tompotika Luwuk University categorized into poor.

Discerning conversational context in online health communities for personalized digital behavior change solutions using Pragmatics to Reveal Intent in Social Media (PRISM) framework

BackgroundOnline health communities (OHCs) have emerged as prominent platforms for behavior modification, and the digitization of online peer interactions has afforded researchers with unique opportunities to model multilevel mechanisms that drive behavior change. Existing studies, however, have been limited by a lack of methods that allow the capture of conversational context and socio-behavioral dynamics at scale, as manifested in these digital platforms. ObjectiveWe develop, evaluate, and apply a novel methodological framework, Pragmatics to Reveal Intent in Social Media (PRISM), to facilitate granular characterization of peer interactions by combining multidimensional facets of human communication. MethodsWe developed and applied PRISM to analyze peer interactions (N = 2.23 million) in QuitNet, an OHC for tobacco cessation. First, we generated a labeled set of peer interactions (n = 2,005) through manual annotation along three dimensions: communication themes (CTs), behavior change techniques (BCTs), and speech acts (SAs). Second, we used deep learning models to apply our qualitative codes at scale. Third, we applied our validated model to perform a retrospective analysis. Finally, using social network analysis (SNA), we portrayed large-scale patterns and relationships among the aforementioned communication dimensions embedded in peer interactions in QuitNet. ResultsQualitative analysis showed that the themes of social support and behavioral progress were common. The most used BCTs were feedback and monitoring and comparison of behavior, and users most commonly expressed their intentions using SAs—expressive and emotion. With additional in-domain pre-training, bidirectional encoder representations from Transformers (BERT) outperformed other deep learning models on the classification tasks. Content-specific SNA revealed that users’ engagement or abstinence status is associated with the prevalence of various categories of BCTs and SAs, which also was evident from the visualization of network structures. ConclusionsOur study describes the interplay of multilevel characteristics of online communication and their association with individual health behaviors.

”That’s Just a Conspiracy Theory!”: Relevant Alternatives, Dismissive Conversational Exercitives, and the Problem of Premature Conclusions

ABSTRACT Drawing on the relevant alternatives framework and Mary Kate McGowan’s work on conversational scorekeeping, I argue that usage of the term ‘conspiracy theory’ in ordinary language and public discourse typically entails the performance of what I call a dismissive conversational exercitive, a kind of speech act that functions to exclude certain propositions from (or prevent their inclusion in) the set of alternatives considered relevant in a given conversational context. While it can be legitimate to perform dismissive conversational exercitives, excluding alternatives that deserve to be taken seriously can be highly problematic for a variety of reasons. For one, it can give rise to what I call the problem of premature conclusions when subjects illegitimately dismiss certain propositions as irrelevant and, as a result, prematurely take certain conclusions or claims to be warranted. Depending on the kind of conclusion or claim, the problem can come in different variants, three of which I shall examine in more detail: the problem of premature knowledge claims, the problem of premature causal claims, and the problem of premature generic generalizations.

Neural response generation for task completion using conversational knowledge graph.

Effective dialogue generation for task completion is challenging to build. The task requires the response generation system to generate the responses consistent with intent and slot values, have diversity in response and be able to handle multiple domains. The response also needs to be context relevant with respect to the previous utterances in the conversation. In this paper, we build six different models containing Bi-directional Long Short Term Memory (Bi-LSTM) and Bidirectional Encoder Representations from Transformers (BERT) based encoders. To effectively generate the correct slot values, we implement a copy mechanism at the decoder side. To capture the conversation context and the current state of the conversation we introduce a simple heuristic to build a conversational knowledge graph. Using this novel algorithm we are able to capture important aspects in a conversation. This conversational knowledge-graph is then used by our response generation model to generate more relevant and consistent responses. Using this knowledge-graph we do not need the entire utterance history, rather only the last utterance to capture the conversational context. We conduct experiments showing the effectiveness of the knowledge-graph in capturing the context and generating good response. We compare these results against hierarchical-encoder-decoder models and show that the use of triples from the conversational knowledge-graph is an effective method to capture context and the user requirement. Using this knowledge-graph we show an average performance gain of 0.75 BLEU score across different models. Similar results also hold true across different manual evaluation metrics.

Speech emotion recognition with artificial intelligence for contact tracing in the COVID‐19 pandemic

AbstractIf understanding sentiments is already a difficult task in human‐human communication, this becomes extremely challenging when a human‐computer interaction happens, as for instance in chatbot conversations. In this work, a machine learning neural network‐based Speech Emotion Recognition system is presented to perform emotion detection in a chatbot virtual assistant whose task was to perform contact tracing during the COVID‐19 pandemic. The system was tested on a novel dataset of audio samples, provided by the company Blu Pantheon, which developed virtual agents capable of autonomously performing contacts tracing for individuals positive to COVID‐19. The dataset provided was unlabelled for the emotions associated to the conversations. Therefore, the work was structured using a sort of transfer learning strategy. First, the model is trained using the labelled and publicly available Italian‐language dataset EMOVO Corpus. The accuracy achieved in testing phase reached 92%. To the best of their knowledge, thiswork represents the first example in the context of chatbot speech emotion recognition for contact tracing, shedding lights towards the importance of the use of such techniques in virtual assistants and chatbot conversational contexts for psychological human status assessment. The code of this work was publicly released at: https://github.com/fp1acm8/SER.

Exploring the Competency of ChatGPT in Solving Competitive Programming Challenges

ChatGPT is an artificial intelligence-powered language processing tool that has gained widespread attention in diverse domains. It uses deep learning techniques to understand conversational context and generate human-like responses. It has also shown its capabilities in the field of computer science and software engineering by generating codes, fixing programming bugs, and explaining programming concepts. This study aims to explore the maturity of ChatGPT by assessing the correctness of solving competitive programming problems. To examine the performance of the responses generated by ChatGPT, numerous competitive programming problems have been chosen covering a variety of topics. In this study, we examine the efficacy of ChatGPT on 300 competitive programming problems spanning different difficulty levels such as Easy, Medium, and Hard. We choose an equal number of problems from each of these three difficulty levels. The performance of problem-solving using ChatGPT shows a promising result compared to the average user. ChatGPT secures an acceptance rate of 89.00%, 68.00%, and 41.00% for Easy, Medium, and Hard difficulty level problems respectively. In the overall assessment, ChatGPT achieves 66.00% acceptance rate whereas the human participants attain 52.95% acceptance rate on average. The results indicate that ChatGPT outperforms the average human problem solvers in solving competitive programming tasks. Moreover, this study also demonstrates the effectiveness of prompts and highlights the limitations of this AI tool and identifies areas where further improvements can be made

Yes, no, maybe so: Caregiver autonomy support, conversation context, and children's memory performance.

The extant literature on the use of autonomy support during caregiver-child conversations has focused primarily on conversations about fun, shared experiences, with limited consideration of unshared experiences or attention toward the role of conversation context. The present study examined how autonomy support, conversation context, and child age interact to predict 3-to-5-year-old children's disclosure of accurate information when discussing an unshared past event with their caregiver and an experimenter. Dyads (N = 111) were recruited from two locations (Miami, Florida and Orange County, California) by research recruitment firms. Children completed a standardized activity alone and then discussed the activity with their caregiver. The context of the discussion was manipulated so that dyads focused on either accumulating facts (Fact condition) or having fun (Fun condition). Afterward, children discussed the activity with a neutral interviewer. Caregivers in the fact condition were less autonomy supportive when discussing the activity than those in the fun condition. During the caregiver-child interview, caregiver autonomy support was negatively associated with children's disclosure of correct event details for those in the fun condition only. Caregiver autonomy support was negatively associated with children's correct details during the experiment-child interview across both context conditions. While older children provided more correct details during both interviews, there were no other age-related effects. These results demonstrate that conversation context moderates the link between autonomy support and children's autobiographical memory performance. Past contradictory findings in the field are discussed in light of these results. (PsycInfo Database Record (c) 2022 APA, all rights reserved).

Encouraging older adults to use assistive devices: The role of politeness theory

ABSTRACT The current study explored the application of Politeness Theory within the context of conversations between older adults and their adult children about acquiring assistive devices. Politeness Theory posits that we want our public self-image, i.e. ‘face’ to be respected, and that concept of face is divided into negative face and positive face. We can show respect for negative face by communicating we value a person’s autonomy, and we can show respect for positive face by communicating we value a person’s abilities. Participants age 65 years or older were randomly assigned to read one of four scripts of a conversation between an older adult and her adult child in which varying levels of respect for autonomy and abilities were expressed. The respect for both autonomy (negative face) and abilities (positive face) group reported significantly higher levels of willingness to consider acquiring assistive devices compared to the threatening both autonomy and ability group. Additionally, the role of geographic regions was explored and the findings were consistent across rural and urban contexts. The use of assistive devices among those who need them has been found to result in pervasive benefits in a variety of areas including physical health, mental health, functional abilities, and social activity. The findings are discussed in terms of implications for education in an array of contexts.

How Would Unconscious Jokes Potentially Influence One's Mindset and Behavior

"Stop rolling, stop rolling" is a kind of self-deprecation and joke that naturally arises in student groups and even adults under the pressure of society's fierce competition. Studies have shown that self-deprecation and jokes have a positive effect on the release of stress, but from daily life experiences and major online platforms, there are also signs of disgust and dislike for such jokes, that is, after the first layer of positive effects, these jokes often have a second layer of subtle and invisible negative effects. According to cognitive psychology, jokes have certain salient and inhibitory effects. For high school students, "Don't roll it" as an unconscious joke can also potentially affect their mindset and behavior. This study found that, first, the relationship between the two people in the conversation constitutes a different conversational context. One of the factors that may make the difference is "the existence of a competitive relationship". That is, when there is a disparity in power between the two people, the other person being joked about tends to be dismissive. Second, when the mental schema of the conversational pair is inconsistent or lacks consensus, the joke may cause the joked student to be offended. Third, when the self-awareness types of the people in this conversation are different, especially when the subject of the joke belongs to a high self-esteem personality type, the joke will hardly work.

Generating Relevant and Informative Questions for Open-Domain Conversations

Recent research has highlighted the importance of mixed-initiative interactions in conversational search. To enable mixed-initiative interactions, information retrieval systems should be able to ask diverse questions, such as information-seeking, clarification, and open-ended ones. question generation (QG) of open-domain conversational systems aims at enhancing the interactiveness and persistence of human-machine interactions. The task is challenging because of the sparsity of question generation (QG)-specific data in conversations. Current work is limited to single-turn interaction scenarios. We propose a context-enhanced neural question generation (CNQG) model that leverages the conversational context to predict question content and pattern, then perform question decoding. A hierarchical encoder framework is employed to obtain the discourse-level context representation. Based on this, we propose Review and Transit mechanisms to respectively select contextual keywords and predict new topic words to further construct the question content. Conversational context and the predicted question content are used to produce the question pattern, which in turn guides the question decoding process implemented by a recurrent decoder with a joint attention mechanism. To fully utilize the limited QG-specific data to train our question generator, we perform multi-task learning with three auxiliary training objectives, i.e., question pattern prediction, Review , and Transit mechanisms. The required additional labeled data is obtained in a self-supervised way. We also design a weight decaying strategy to adjust the influences of various auxiliary learning tasks. To the best of our acknowledge, we are the first to extend the application of QG to the multi-turn open-domain conversational scenario. Extensive experimental results demonstrate the effectiveness of our proposal and its main components on generating relevant and informative questions, with robust performance for contexts with various lengths.

Verbal anthropomorphism design of social robots: Investigating users’ privacy perception

Many studies on the anthropomorphism of social robots have revealed two aspects of anthropomorphism in human–robot interaction. The first is that humanlike robots have advantages of emotional, functional, and relational interaction with users, and the second is that they have disadvantages of privacy perception. Some researchers have claimed that there would be an optimal anthropomorphism of social robots balancing the two aspects. In this study, we explore the optimal level of verbal anthropomorphism of social robots through privacy perception. We designed social robots’ verbal anthropomorphism with various verbal style cues and conducted two studies related to physical and informational privacy based on three levels of verbal anthropomorphism: high, medium, and low. The results revealed that a medium-level verbal anthropomorphism was preferred by participants in terms of privacy perception, whereas they preferred a high-level verbal anthropomorphism in terms of emotional, functional, and relational aspects. This implies that the optimal level of verbal anthropomorphism of social robots may vary depending on privacy perceptions. Based on these findings, we propose adaptable verbal anthropomorphism design of social robots depending on conversation contexts. This paper offers directions to researchers and practitioners who want to design the verbal anthropomorphism of social robots.