Quality Of Answers Research Articles

ChatGPT as a Source for Patient Information on Patellofemoral Surgery—A Comparative Study Amongst Laymen, Doctors, and Experts

Introduction: In November 2022, OpenAI launched ChatGPT for public use through a free online platform. ChatGPT is an artificial intelligence (AI) chatbot trained on a broad dataset encompassing a wide range of topics, including medical literature. The usability in the medical field and the quality of AI-generated responses are widely discussed and are the subject of current investigations. Patellofemoral pain is one of the most common conditions among young adults, often prompting patients to seek advice. This study examines the quality of ChatGPT as a source of information regarding patellofemoral conditions and surgery, hypothesizing that there will be differences in the evaluation of responses generated by ChatGPT between populations with different levels of expertise in patellofemoral disorders. Methods: A comparison was conducted between laymen, doctors (non-orthopedic), and experts in patellofemoral disorders based on a list of 12 questions. These questions were divided into descriptive and recommendatory categories, with each category further split into basic and advanced content. Questions were used to prompt ChatGPT in April 2024 using the ChatGPT 4.0 engine, and answers were evaluated using a custom tool inspired by the Ensuring Quality Information for Patients (EQIP) instrument. Evaluations were performed independently by laymen, non-orthopedic doctors, and experts, with the results statistically analyzed using a Mann–Whitney U Test. A p-value of less than 0.05 was considered statistically significant. Results: The study included data from seventeen participants: four experts in patellofemoral disorders, seven non-orthopedic doctors, and six laymen. Experts rated the answers lower on average compared to non-experts. Significant differences were observed in the ratings of descriptive answers with increasing complexity. The average score for experts was 29.3 ± 5.8, whereas non-experts averaged 35.3 ± 5.7. For recommendatory answers, experts also gave lower ratings, particularly for more complex questions. Conclusion: ChatGPT provides good quality answers to questions concerning patellofemoral disorders, although questions with higher complexity were rated lower by patellofemoral experts compared to non-experts. This study emphasizes the potential of ChatGPT as a complementary tool for patient information on patellofemoral disorders, although the quality of the answers fluctuates with the complexity of the questions, which might not be recognized by non-experts. The lack of personalized recommendations and the problem of “AI hallucinations” remain a challenge. Human expertise and judgement, especially from trained healthcare experts, remain irreplaceable.

Специфика усвоения диалоговых навыков дошкольниками с расстройствами аутистического спектра

<p style="text-align: justify;"><strong>Objectives. </strong>The variability of disorders in the formation of dialog skills of preschoolers with autism spectrum disorders (ASD) does not always depend on the level of intellectual development. In issues of speech development, it is important to take into account both the specifics of understanding the reversed speech at the level of interrogative statements, and the aspects of structuring and semantics of the answer proposed by the child with ASD himself. In order to study the specifics of mastering the basic conversational skills by children with ASD who have unimpaired intellect, a special pilot study was designed and conducted.</p> <p style="text-align: justify;"><strong>Methods. </strong>The peculiarities of understanding questions of various purposes and structures by students of the age of 5—7 years (group 1 — 51 children) with unimpaired intellect and ability of verbal communication were studied. Children of group 1 and preschoolers of group 2 of younger age (from 3 years 10 months to 5 years 2 months, a total of 28 participants) were presented with 48 questions, hierarchically organized based on ontogenetic patterns associated with the formation of cognitive concepts, which are accessible to a typically developing child by a certain age. The results obtained in both groups were compared.</p> <p style="text-align: justify;"><strong>Results. </strong>Only 2.3% of the proposed 3,708 questions remained unanswered, and the number of correct (adequate) answers, depending on age, tended to increase (for example, preschoolers with ASD of 4 years gave 68% correct answers, preschoolers of 5 years — 73%). The study shows that, in terms of nature and specifics of dialog skills, students with ASD 5-7 years old have no significant differences from the typically developing preschoolers of a younger age group: the number and quality of correct answers from children with ASD at the age of 6 years is comparable with the number of correct answers from typically developing preschoolers of 4 years old.</p> <p style="text-align: justify;"><strong>Conclusions. </strong>In the process of forming dialog skills in the structure of developmental education for children with ASD, the teacher should rely solely on the traditional spontaneous question-and-answer method, they should also consider the semantics and pragmatics of the proposed lexical material. The data obtained will be useful for special education teachers in selecting the content for speech development, taking into account the individual typological characteristics of children with ASD.</p>

Caching Historical Embeddings in Conversational Search

Rapid response, namely, low latency, is fundamental in search applications; it is particularly so in interactive search sessions, such as those encountered in conversational settings. An observation with a potential to reduce latency asserts that conversational queries exhibit a temporal locality in the lists of documents retrieved. Motivated by this observation, we propose and evaluate a client-side document embedding cache, improving the responsiveness of conversational search systems. By leveraging state-of-the-art dense retrieval models to abstract document and query semantics, we cache the embeddings of documents retrieved for a topic introduced in the conversation, as they are likely relevant to successive queries. Our document embedding cache implements an efficient metric index, answering nearest-neighbor similarity queries by estimating the approximate result sets returned. We demonstrate the efficiency achieved using our cache via reproducible experiments based on Text Retrieval Conference Conversational Assistant Track datasets, achieving a hit rate of up to 75% without degrading answer quality. Our achieved high cache hit rates significantly improve the responsiveness of conversational systems while likewise reducing the number of queries managed on the search back-end.

Using ChatGPT to Provide Patient-Specific Answers to Parental Questions in the PICU.

To determine if ChatGPT can incorporate patient-specific information to provide high-quality answers to parental questions in the PICU. We hypothesized that ChatGPT would generate high-quality, patient-specific responses. In this cross-sectional study, we generated assessments and plans for 3 PICU patients with respiratory failure, septic shock, and status epilepticus and paired them with 8 typical parental questions. We prompted ChatGPT with instructions, an assessment and plan, and 1 question. Six PICU physicians evaluated the responses for accuracy (1-6), completeness (yes/no), empathy (1-6), and understandability (Patient Education Materials Assessment Tool, PEMAT, 0% to 100%; Flesch-Kincaid grade level). We compared answer quality among scenarios and question types using the Kruskal-Wallis and Fischer's exact tests. We used percent agreement, Cohen's Kappa, and Gwet's agreement coefficient to estimate inter-rater reliability. All answers incorporated patient details, utilizing them for reasoning in 59% of sentences. Responses had high accuracy (median 5.0, [interquartile range (IQR), 4.0-6.0]), empathy (median 5.0, [IQR, 5.0-6.0]), completeness (97% of all questions), and understandability (PEMAT % median 100, [IQR, 87.5-100]; Flesch-Kincaid level 8.7). Only 4/144 reviewer scores were <4/6 in accuracy, and no response was deemed likely to cause harm. There was no difference in accuracy, completeness, empathy, or understandability among scenarios or question types. We found fair, substantial, and almost perfect agreement among reviewers for accuracy, empathy, and understandability, respectively. ChatGPT used patient-specific information to provide high-quality answers to parental questions in PICU clinical scenarios.

What motivates knowledge sharing? Evaluating the quality of answer contribution in online Q&A communities

The Online Q&A community provides platforms for Internet users to exchange and share knowledge. Its rapid development has caused the problem of information overload, which promotes users' demand for precise and personalized information. In light of this, we innovatively analyze the correlation between knowledge quality and user behaviors. First, we propose a novel approach method to evaluate the answer quality through a machine learning approach, which applies the information adoption theory to the design of a text classification model. Hereafter, with applications of motivation crowding theory, this paper introduces the classification results of answer quality by machine learning algorithms into the empirical research model to explore the factors that motivate users in both participation and high-quality user-generated content creation. Results show that both extrinsic and intrinsic factors determine the quality of knowledge contributors' answers. Further, certain external interventions (monetary rewards) can crowd out the effects of knowledge contributors’ intrinsic motivations (knowledge self-efficacy). It enriches the research on user knowledge contribution behavior in online Q&A communities and also provides theoretical guidance and suggestions for community operation.

Application of a Question Answering System in the Ontology-Based Health Domain and Question Templates to Improve the Quality of Answers

Developing an effective Question Answering System (QAS) to address a wide range of health-related inquiries presents a significant challenge. This study introduces an innovative approach using an ontology-based QAS within the health domain, leveraging the Resource Description Framework (RDF) and SPARQL query language. By utilizing ontology, the system can more accurately map health concepts, improving the relevance and precision of its responses. Additionally, the incorporation of structured Question Templates enhances the system’s ability to understand and respond to diverse user queries. The system's performance was evaluated using Black Box Testing, which demonstrated that it consistently provides accurate and relevant answers. The system achieved a Mean Average Precision (MAP) of 79.6%, indicating its potential to effectively address a broad spectrum of health issues.

Exploring the Efficacy of a ChatGPT-Based Application in Enhancing Oral English Proficiency for Chinese ESL Students

This research explores the impact of Chat GPT AI Characters on ESL learning, comparing them to traditional peer practice in language education. It investigates AI's influence on language proficiency, engagement, and response quality, drawing from literature on ESL challenges. Using mixed methods—pre- and post-test questionnaires, classroom experiments—the study evaluates two groups' language fluency, response speed, and answer quality. Thematic analysis uncovers student perspectives, while statistical methods like t-tests and correlation studies highlight group differences. Ethical considerations prioritize privacy and consent. Anticipated outcomes include improved language skills and engagement. Acknowledging limitations in sample size and cultural biases, this study illuminates AI's role in ESL education, informing future teaching practices.

Deciphering digital popularity from question quality: the preeminence of social attributes over textual dimensions in driving CQA platform traffic

ABSTRACT Community Question Answering (CQA) sites, now navigating through millions of daily user visits, hinge significantly on the quality of user-generated questions to bolster user satisfaction and, consequently, website traffic – crucial for platform sustainability. While the quality of questions ostensibly sets the stage for insightful answers, thereby attracting substantial traffic, existing research has predominantly sidelined the significance of questions, favoring a focus on traditional factors, such as answer quality, to elucidate traffic determinants. In this study we quantified the interrelation between question quality and traffic by extracting 15 relevant question-quality features and corresponding traffic (i.e. pageview) data from a leading Chinese CQA platform. Utilizing Grey Relational Analysis, we assessed these attributes’ relative influence on pageviews and created a ranking list of features of questions. Based on the rankings, our findings spotlight the paramount influence of a question's digital popularity and attributes tied to the questioner – central pillars of social features – surpassing the impact of textual elements. This study further delineates the underlying reasons for such patterns based on traffic sourced from diverse channels. Our revelations offer pivotal insights for optimizing CQA platform engagement and furnish avenues for prospective research.

Still Using Only ChatGPT? The Comparison of Five Different Artificial Intelligence Chatbots' Answers to the Most Common Questions About Kidney Stones.

Objective: To evaluate and compare the quality and comprehensibility of answers produced by five distinct artificial intelligence (AI) chatbots-GPT-4, Claude, Mistral, Google PaLM, and Grok-in response to the most frequently searched questions about kidney stones (KS). Materials and Methods: Google Trends facilitated the identification of pertinent terms related to KS. Each AI chatbot was provided with a unique sequence of 25 commonly searched phrases as input. The responses were assessed using DISCERN, the Patient Education Materials Assessment Tool for Printable Materials (PEMAT-P), the Flesch-Kincaid Grade Level (FKGL), and the Flesch-Kincaid Reading Ease (FKRE) criteria. Results: The three most frequently searched terms were "stone in kidney," "kidney stone pain," and "kidney pain." Nepal, India, and Trinidad and Tobago were the countries that performed the most searches in KS. None of the AI chatbots attained the requisite level of comprehensibility. Grok demonstrated the highest FKRE (55.6 ± 7.1) and lowest FKGL (10.0 ± 1.1) ratings (p = 0.001), whereas Claude outperformed the other chatbots in its DISCERN scores (47.6 ± 1.2) (p = 0.001). PEMAT-P understandability was the lowest in GPT-4 (53.2 ± 2.0), and actionability was the highest in Claude (61.8 ± 3.5) (p = 0.001). Conclusion: GPT-4 had the most complex language structure of the five chatbots, making it the most difficult to read and comprehend, whereas Grok was the simplest. Claude had the best KS text quality. Chatbot technology can improve healthcare material and make it easier to grasp.

Evaluating ChatGPT as a patient resource for frequently asked questions about lung cancer surgery—a pilot study

ObjectiveChat-based artificial intelligence programs like ChatGPT are reimagining how patients seek information. This study aims to evaluate the quality and accuracy of ChatGPT-generated answers to common patient questions about lung cancer surgery. MethodsA 30-question survey of patient questions about lung cancer surgery was posed to ChatGPT in July 2023. The ChatGPT-generated responses were presented to 9 thoracic surgeons at 4 academic institutions who rated the quality of the answer on a 5-point Likert scale. They also evaluated if the response contained any inaccuracies and were prompted to submit free text comments. Responses were analyzed in aggregate. ResultsFor ChatGPT-generated answers, the average quality ranged from 3.1 to 4.2 of 5.0, indicating they were generally “good” or “very good.” No answer received a unanimous 1-star (poor quality) or 5-star (excellent quality) score. Minor inaccuracies were found by at least 1 surgeon in 100% of the answers, and major inaccuracies were found in 36.6%. Regarding ChatGPT, 66.7% of surgeons thought it was an accurate source of information for patients. However, only 55.6% thought they were comparable with answers given by experienced thoracic surgeons, and only 44.4% would recommend it to their patients. Common criticisms of ChatGPT-generated answers included lengthiness, lack of specificity regarding surgical care, and lack of references. ConclusionsChat-based artificial intelligence programs have potential to become a useful information tool for patients with lung cancer surgery. However, the quality and accuracy of ChatGPT-generated answers need improvement before thoracic surgeons would consider this method as a primary education source for patients.

Accessing information provided via artificial intelligence regarding reverse and anatomic total shoulder arthroplasty

BackgroundThe purpose of this study is to analyze the quality, accuracy, and readability of information provided by an artificial intelligence (AI) interface ChatGPT (OpenAI, San Francisco). We searched ChatGPT for commonly asked questions by patients regarding anatomic total shoulder arthroplasty (aTSA) and reverse total shoulder arthroplasty (rTSA). MethodsChatGPT was used to answer 30 commonly asked questions by patients regarding aTSA and rTSA, inputted as “total shoulder replacement” and “reverse shoulder replacement”. These questions were categorized based on the Rothwell criteria into Fact, Policy, and Value. The answers generated were analyzed for quality, accuracy, and readability using the DISCERN scale, JAMA benchmark criteria, Flesch-Kincaid Reading Ease Score (FRES) and Grade Level (FKGL). ResultsFor both rTSA and aTSA the DISCERN score for Fact questions was 57, Policy questions was 61, and for Value questions was 58 (all were all considered “good”). The JAMA benchmark criteria was 0, representing the lowest score for Fact, Policy, and Value questions for both rTSA and aTSA questions. The FRES score for the aTSA answers for Fact was 15.15, for Policy was 11.14, and for Value questions was 10.95. The FRES score for rTSA questions for Fact is 48.02, Policy is 12.51, and Value is 17.22. The FKGL for aTSA answer for Fact was 17.48, Policy was 17.72 and Value was 17.96. The FKGL for rTSA questions for Fact are 8.10, Policy is 17.27, and Value is 16.56. ConclusionOverall, the quality of answers provided by AI open model, ChatGPT was considered “good.” The information provided had lower reliability, and lack of information regarding funding and disclosures. Most of the information generated by ChatGPT was also found to have the readability of “academic level text”, while Fact related information on reverse shoulder arthroplasty was found to have the readability of 9th grade level, which may be too complex for most patients. Overall, these results indicate that ChatGPT can provide correct answers to questions about aTSA and rTSA, although we would caution patients from utilizing this resource due to the lack of citations and complexity of the output that ChatGPT provides. Importantly, all answers provided by AI suggested reaching out to physicians to get more accurate and personalized advise, to factor into the shared decisions making model.

Optimizing biomedical information retrieval with a keyword frequency-driven prompt enhancement strategy

BackgroundMining the vast pool of biomedical literature to extract accurate responses and relevant references is challenging due to the domain's interdisciplinary nature, specialized jargon, and continuous evolution. Early natural language processing (NLP) approaches often led to incorrect answers as they failed to comprehend the nuances of natural language. However, transformer models have significantly advanced the field by enabling the creation of large language models (LLMs), enhancing question-answering (QA) tasks. Despite these advances, current LLM-based solutions for specialized domains like biology and biomedicine still struggle to generate up-to-date responses while avoiding “hallucination” or generating plausible but factually incorrect responses.ResultsOur work focuses on enhancing prompts using a retrieval-augmented architecture to guide LLMs in generating meaningful responses for biomedical QA tasks. We evaluated two approaches: one relying on text embedding and vector similarity in a high-dimensional space, and our proposed method, which uses explicit signals in user queries to extract meaningful contexts. For robust evaluation, we tested these methods on 50 specific and challenging questions from diverse biomedical topics, comparing their performance against a baseline model, BM25. Retrieval performance of our method was significantly better than others, achieving a median Precision@10 of 0.95, which indicates the fraction of the top 10 retrieved chunks that are relevant. We used GPT-4, OpenAI's most advanced LLM to maximize the answer quality and manually accessed LLM-generated responses. Our method achieved a median answer quality score of 2.5, surpassing both the baseline model and the text embedding-based approach. We developed a QA bot, WeiseEule (https://github.com/wasimaftab/WeiseEule-LocalHost), which utilizes these methods for comparative analysis and also offers advanced features for review writing and identifying relevant articles for citation.ConclusionsOur findings highlight the importance of prompt enhancement methods that utilize explicit signals in user queries over traditional text embedding-based approaches to improve LLM-generated responses for specialized queries in specialized domains such as biology and biomedicine. By providing users complete control over the information fed into the LLM, our approach addresses some of the major drawbacks of existing web-based chatbots and LLM-based QA systems, including hallucinations and the generation of irrelevant or outdated responses.

BioASQ Synergy: a dialogue between question-answering systems and biomedical experts for promoting COVID-19 research.

This article presents the novel BioASQ Synergy research process which aims to facilitate the interaction between biomedical experts and automated question-answering systems. The proposed research allows systems to provide answers to emerging questions, which in turn are assessed by experts. The assessment of the experts is fed back to the systems, together with new questions. With this iteration, we aim to facilitate the incremental understanding of a developing problem and contribute to solution discovery. The results suggest that the proposed approach can assist researchers to navigate available resources. The experts seem to be very satisfied with the quality of the ideal answers provided by the systems, suggesting that such systems are already useful in answering open research questions. BioASQ Synergy aspires to provide a tool that gives the experts easy and personalized access to the latest findings in a fast-growing corpus of material. In this article, we envisioned BioASQ Synergy as a continuous dialogue between experts and systems to issue open questions. We ran an initial proof-of-concept of the approach, in order to evaluate its usefulness, both from the side of the experts, as well as from the side of the participating systems.

Analyzing the performance of ChatGPT in answering inquiries about cervical cancer.

To analyze the knowledge of ChatGPT about cervical cancer (CC). Official websites of professional health institutes, and websites created by patients and charities underwent strict screening. Using CC-related keywords, common inquiries by the public and comments about CC were searched in social media applications with these data, a list of frequently asked questions (FAQs) was prepared. When preparing question about CC, the European Society of Gynecological Oncology (ESGO), European Society for Radiotherapy and Oncology (ESTRO), and European Society of Pathology (ESP) guidelines were used. The answers given by ChatGPT were scored according to the Global Quality Score (GQS). When all ChatGPT answers to FAQs about CC were evaluated with regard to GQS, 68 ChatGPT answers were classified as score 5, and none of ChatGPT answers for FAQs were scored as 2 or 1. Moreover, ChatGPT answered 33 of 53 (62.3%) CC-related questions based on ESGO, ESTRO, and ESP guidelines with completely accurate and satisfactory responses (GQS 5). In addition, eight answers (15.1%), seven answers (13.2%), four answers (7.5%), and one answer (1.9%) were categorized as GQS 4, GQS 3, GQS 2, and GQS 1, respectively. The reproducibility rate of ChatGPT answers about CC-related FAQs and responses about those guideline-based questions was 93.2% and 88.7%, respectively. ChatGPT had an accurate and satisfactory response rate for FAQs about CC with regards to GQS. However, the accuracy and quality of ChatGPT answers significantly decreased for questions based on guidelines.

Evaluation of Artificial Intelligence as a Search Tool for Patients: Can ChatGPT-4 Provide Accurate Evidence-Based Orthodontic-Related Information?

Artificial intelligence (AI) is already a part of our reality. Many people started using ChatGPT in their daily life, replacing existing web browsers. The confidence people put in the ability of ChatGPT to provide accurate medical information is increasing. With that, the need for proper assessment tools for the safety and reliability of ChatGPT is also crucial. This study aimed to assess the accuracy, reliability, andquality of information provided by ChatGPT-4 on three specific orthodontic topics, namely, impacted canines, interceptive orthodontic treatment, and orthognathic surgery, as evaluated by five experienced orthodontists using a Likert scale ranking method. Using ChatGPT version 4, 20 most commonly asked questions were generated and answered on the following topics: impacted canines, interceptive treatment, and orthognathic surgery. The evaluation of the quality of the answers provided was done by five experienced orthodontists. Quality assessment was done using the Likert scale ranking method. The quality answers generated by a conversational AI system (ChatGPT4) were evaluated by five experienced orthodontists for three topics: impacted canines, interceptive orthodontics, and orthognathic surgery. The evaluators rated each question-answer pair on a five-point scale from "very poor" to "very good." The results showed that the AI system produced generally good quality information for all topics, with no significant difference between them. The inter-rater agreement among the experts was low, indicating some variability in their judgments. This study demonstrates that ChatGPT4 can provide generally good information on impacted canines, interceptive treatment, and orthognathic surgery. However, answers provided should be handled with caution due to variability and lack of reliability and should not be considered a substitute for professional opinion.

On the Use of Generative AI for Literature Reviews: An Exploration of Tools and Techniques

To carry out a literature review often involves hard and tedious work. There is a tradition of using facilitating tools, that extended to the AI field in 2018 when iris.ai appeared. Today, in the emerging field of Generative AI tools based on Large Language Models, there has been rapid development of new literature search tools and approaches. This study has the aim of exploring this vast array of Generative AI tools, in a literature study where some of the found tools were used to facilitate the selection of relevant publication. Three research questions guided this study: RQ1) "What Generative AI tools can be found in literature?", RQ2) "Which of these tools could be of use in the literature review conducted in this study, and how?", and RQ3) "What are the ethical aspects of using Generative AI tools in literature studies?” The approach has been a scoping review, built around a search that combined the keywords: "AI supported", "AI generated", "AI based" and "Literature review". An initial result set was filtered with inclusion exclusion criteria in a strive for an interesting quality answer to the research questions. However, most publications that passed the filtering lacked any potential to contribute to answer the research questions. The most interesting finding in the first search was a hint about the new feature 'Scopus AI'. A new search with the Scopus AI tool resulted in a small but very relevant set of publications. These publications were analysed in a deductive inductive thematic analysis, and primary sorted into the categories of: 'Generative AI Tools', 'Supportive AI Techniques', and 'Ethical Issues'. Findings indicate that there is a wide variety of tools that can facilitate the skimming process of a literature, and to provide adequate summaries of retrieved publication. However, authors recommendation is to keep the tools on the facilitating support level, and that the main analysis and conclusion should be human conducted. With this, rather traditional approach, researchers will have clearly less ethical issues to consider. Finally, the ethical aspects of Generative AI tools in research ought to be investigated more in detail, in a separate future study.

Mindreading quality versus quantity: A theoretically and empirically motivated two-factor structure for individual differences in adults' mindreading.

Existing methods for studying individual differences in adults' mindreading often lack good psychometric characteristics. Moreover, it remains unclear, even in theory, how mindreading varies in adults who already possess an understanding of mental states. In this pre-registered study, it was hypothesised that adults vary in their motivation for mindreading and in the degree to which their answers on mindreading tasks are appropriate (context-sensitive). These factors are confounded in existing measures as they do not differentiate between the frequency of mental state terms (MST), indicative of motivation, and the quality of an explanation. Using an innovative scoring system, the current study examined whether individual differences in adult undergraduate psychology students' (N = 128) answer quality and / or quantity of explicit references to others' mental states on two open-ended response mindreading tasks were separable constructs, accounted for by mindreading motivation, and related differentially to measures previously linked with mindreading (e.g., religiosity, loneliness, social network size). A two-factor and one-factor model both provided acceptable fit. Neither model showed significant associations with mindreading motivation. However, a two-factor model (with MST and response appropriateness loading onto separate factors) provided greater explanatory power. Specifically, MST was positively associated with religiosity and response appropriateness was negatively associated with religiosity, whilst the one-factor solution did not predict any socially relevant outcomes. This provides some indication that mindreading quantity and mindreading quality may be distinguishable constructs in the structure of individual differences in mindreading.

Blockchain-based Crowdsourcing for Human Intelligence Tasks with Dual Fairness

Human intelligence tasks (HITs) are widely utilized for crowdsourcing human knowledge, such as labeling images for machine learning. Centralized crowdsourcing platforms face challenges of a single point of failure and a lack of service transparency. Existing blockchain-based crowdsourcing approaches overlook the low scalability problem of permissionless blockchains or inconveniently rely on existing ground-truth data as the root of trust to evaluate quality of workers' answers. We propose a blockchain-based crowdsourcing scheme for ensuring dual fairness (i.e., preventing false-reporting and free-riding) and improving on-chain efficiency concerning on-chain storage and smart contract computation. The proposed scheme does not rely on trusted authorities but rather depends on a public blockchain to guarantee the dual fairness. An efficient and publicly verifiable truth discovery scheme is designed based on majority voting and cryptographic accumulators. This truth discovery scheme aims at inferring ground truth from workers' answers. The ground truth is further utilized to estimate the quality of workers' answers. Additionally, a novel blockchain-based protocol is designed to further reduce on-chain costs while ensuring truthfulness. The scheme has O(n) complexity for both on-chain storage and smart contract computation, regardless of the number of questions, where n denotes the number of workers. Formal security analysis is provided, and extensive experiments are conducted to evaluate effectiveness and performance.

Penerapan Langchain Retriever dengan Model Chat Openai dalam Pengembangan Sistem Chatbot Hadis Berbasis Telegram

In Islamic studies, the Hadiths of Prophet Muhammad (SAW) hold significant value as guides for behavior and faith. However, access to understanding Hadiths often presents challenges, espe-cially for those who are not Hadith experts. The digitalization of Hadiths is still limited, making it time-consuming to find answers by sifting through the vast amount of available information. This research aims to create an efficient chatbot that provides answers related to Hadiths, including the original sources, quickly. The proposed solution is a technology-based approach through the development of a Hadith chatbot on Telegram, integrated with the LangChain Retriever and the GPT-4-1106-preview chat model from OpenAI. Using LangChain Retriever helps the chatbot find accurate answers by matching user questions with relevant Hadith databases, enhancing the ac-curacy of the chatbot's responses. The GPT-4-1106-preview chat model enables the chatbot to generate natural and context-appropriate responses, improving user interaction. The Rapid Ap-plication Development (RAD) method is applied in system development, through stages of Re-quirement Planning, User Design, Construction, and Cut-Over, including data analysis of Hadiths from the Nine Imam Hadith Books, totaling 62,169 Hadiths. The chatbot's performance evaluation uses the Scoring Evaluator framework with an average evaluation score of 0.97 and quality answer evaluation testing by five Hadith experts with an accuracy percentage of 90%. The Scoring Eval-uator test results indicate that the responses are highly accurate and aligned with Hadith refer-ences, and the quality answer evaluation test on a Likert scale shows respondents strongly agree with the system's answers. This research contributes to laypersons wanting to learn Hadiths by utilizing the chatbot as an interactive and innovative learning medium. Further research can expand the focus to complex interpretations of Musykil al-Hadith and asbab al-wurud to address deeper questions about Hadith interpretation.

“What Scares Me Is the Speed at Which Artificial Intelligence Is Developing”: Students' Perceptions of Artificial Intelligence in Foreign Language Teaching

As artificial intelligence (AI) becomes an integral part of our daily lives, the concern of the teaching community about the illegal use of these technologies in the educational process is increasing. In order to adapt the education system and teaching practices to new technological challenges, it is necessary to analyze the opinions of all the parties concerned. The purpose of this study is to identify the attitude of students of Kazan Federal University to the use of artificial intelligence technologies in the educational process and the practice of their application in foreign languages learning process. To achieve this goal, an online survey of students of Kazan Federal University was conducted. The survey touched upon the practical aspects of the use of artificial intelligence in language teaching, the advantages and disadvantages of AI tools from the students’ point of view, as well as their opinion regarding the prospects of AI in education. As a result of the study, we came to the conclusion that at the moment AI tools are not widespread enough in teaching foreign languages. Only one-fifth of the respondents use these tools, but the respondents’ comments suggest that the number of users will grow. The attitude of students towards the use of AI is ambiguous, with responses ranging from enthusiastic to skeptical. Students’ positive impressions are mainly related to saving time and effort, as well as to the ability of AI to present complex materials with understandable language. Among the main disadvantages, the respondents noted unreliability of data and fake content. Despite the fact that students are generally positive about the use of AI, a significant part of respondents do not trust software products such as ChatGPT, since, in their opinion, it provides answers of average quality that need to be corrected. Based on the data obtained, the authors attempted to formulate recommendations on improving the methods of teaching and control in the process of teaching foreign languages at universities.