arXiv - CS - Computers and Society最新文献

{"title":"Including Non-Autistic Peers in Games Designed for Autistic Socialization","authors":"Yiqi Xiao","doi":"arxiv-2408.13967","DOIUrl":"https://doi.org/arxiv-2408.13967","url":null,"abstract":"Through a review of current game practices, the author highlights concerns\u0000regarding the safety of public social games and the singular medical approach\u0000to serious game design for autism. The paper identifies a disconnect between\u0000the needs of autistic children and the existing solutions. To fill this gap, a\u0000neurodiversity approach to serious game design is proposed. This approach aims\u0000to address the social needs of autistic children, enabling them to interact\u0000with their neurotypical peers directly, confidently, and safely.","PeriodicalId":501112,"journal":{"name":"arXiv - CS - Computers and Society","volume":"11 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142183923","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Reconciling Different Theories of Learning with an Agent-based Model of Procedural Learning","authors":"Sina Rismanchian, Shayan Doroudi","doi":"arxiv-2408.13364","DOIUrl":"https://doi.org/arxiv-2408.13364","url":null,"abstract":"Computational models of human learning can play a significant role in\u0000enhancing our knowledge about nuances in theoretical and qualitative learning\u0000theories and frameworks. There are many existing frameworks in educational\u0000settings that have shown to be verified using empirical studies, but at times\u0000we find these theories make conflicting claims or recommendations for\u0000instruction. In this study, we propose a new computational model of human\u0000learning, Procedural ABICAP, that reconciles the ICAP,\u0000Knowledge-Learning-Instruction (KLI), and cognitive load theory (CLT)\u0000frameworks for learning procedural knowledge. ICAP assumes that constructive\u0000learning generally yields better learning outcomes, while theories such as KLI\u0000and CLT claim that this is not always true. We suppose that one reason for this\u0000may be that ICAP is primarily used for conceptual learning and is\u0000underspecified as a framework for thinking about procedural learning. We show\u0000how our computational model, both by design and through simulations, can be\u0000used to reconcile different results in the literature. More generally, we\u0000position our computational model as an executable theory of learning that can\u0000be used to simulate various educational settings.","PeriodicalId":501112,"journal":{"name":"arXiv - CS - Computers and Society","volume":"4 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142183925","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Quantitative Information Extraction from Humanitarian Documents","authors":"Daniele Liberatore, Kyriaki Kalimeri, Derya Sever, Yelena Mejova","doi":"arxiv-2408.04941","DOIUrl":"https://doi.org/arxiv-2408.04941","url":null,"abstract":"Humanitarian action is accompanied by a mass of reports, summaries, news, and\u0000other documents. To guide its activities, important information must be quickly\u0000extracted from such free-text resources. Quantities, such as the number of\u0000people affected, amount of aid distributed, or the extent of infrastructure\u0000damage, are central to emergency response and anticipatory action. In this\u0000work, we contribute an annotated dataset for the humanitarian domain for the\u0000extraction of such quantitative information, along side its important context,\u0000including units it refers to, any modifiers, and the relevant event. Further,\u0000we develop a custom Natural Language Processing pipeline to extract the\u0000quantities alongside their units, and evaluate it in comparison to baseline and\u0000recent literature. The proposed model achieves a consistent improvement in the\u0000performance, especially in the documents pertaining to the Dominican Republic\u0000and select African countries. We make the dataset and code available to the\u0000research community to continue the improvement of NLP tools for the\u0000humanitarian domain.","PeriodicalId":501112,"journal":{"name":"arXiv - CS - Computers and Society","volume":"15 Suppl 1 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141930919","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Examining the Behavior of LLM Architectures Within the Framework of Standardized National Exams in Brazil","authors":"Marcelo Sartori Locatelli, Matheus Prado Miranda, Igor Joaquim da Silva Costa, Matheus Torres Prates, Victor Thomé, Mateus Zaparoli Monteiro, Tomas Lacerda, Adriana Pagano, Eduardo Rios Neto, Wagner Meira Jr., Virgilio Almeida","doi":"arxiv-2408.05035","DOIUrl":"https://doi.org/arxiv-2408.05035","url":null,"abstract":"The Exame Nacional do Ensino M'edio (ENEM) is a pivotal test for Brazilian\u0000students, required for admission to a significant number of universities in\u0000Brazil. The test consists of four objective high-school level tests on Math,\u0000Humanities, Natural Sciences and Languages, and one writing essay. Students'\u0000answers to the test and to the accompanying socioeconomic status questionnaire\u0000are made public every year (albeit anonymized) due to transparency policies\u0000from the Brazilian Government. In the context of large language models (LLMs),\u0000these data lend themselves nicely to comparing different groups of humans with\u0000AI, as we can have access to human and machine answer distributions. We\u0000leverage these characteristics of the ENEM dataset and compare GPT-3.5 and 4,\u0000and MariTalk, a model trained using Portuguese data, to humans, aiming to\u0000ascertain how their answers relate to real societal groups and what that may\u0000reveal about the model biases. We divide the human groups by using\u0000socioeconomic status (SES), and compare their answer distribution with LLMs for\u0000each question and for the essay. We find no significant biases when comparing\u0000LLM performance to humans on the multiple-choice Brazilian Portuguese tests, as\u0000the distance between model and human answers is mostly determined by the human\u0000accuracy. A similar conclusion is found by looking at the generated text as,\u0000when analyzing the essays, we observe that human and LLM essays differ in a few\u0000key factors, one being the choice of words where model essays were easily\u0000separable from human ones. The texts also differ syntactically, with LLM\u0000generated essays exhibiting, on average, smaller sentences and less thought\u0000units, among other differences. These results suggest that, for Brazilian\u0000Portuguese in the ENEM context, LLM outputs represent no group of humans, being\u0000significantly different from the answers from Brazilian students across all\u0000tests.","PeriodicalId":501112,"journal":{"name":"arXiv - CS - Computers and Society","volume":"107 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141930917","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Criticizing Ethics According to Artificial Intelligence","authors":"Irina Spiegel","doi":"arxiv-2408.04609","DOIUrl":"https://doi.org/arxiv-2408.04609","url":null,"abstract":"This article presents a critique of ethics in the context of artificial\u0000intelligence (AI). It argues for the need to question established patterns of\u0000thought and traditional authorities, including core concepts such as autonomy,\u0000morality, and ethics. These concepts are increasingly inadequate to deal with\u0000the complexities introduced by emerging AI and autonomous agents. This critique\u0000has several key components: clarifying conceptual ambiguities, honestly\u0000addressing epistemic issues, and thoroughly exploring fundamental normative\u0000problems. The ultimate goal is to reevaluate and possibly redefine some\u0000traditional ethical concepts to better address the challenges posed by AI.","PeriodicalId":501112,"journal":{"name":"arXiv - CS - Computers and Society","volume":"76 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141930920","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Towards the Socio-Algorithmic Construction of Fairness: The Case of Automatic Price-Surging in Ride-Hailing","authors":"Mateusz Dolata, Gerhard Schwabe","doi":"arxiv-2408.04685","DOIUrl":"https://doi.org/arxiv-2408.04685","url":null,"abstract":"Algorithms take decisions that affect humans, and have been shown to\u0000perpetuate biases and discrimination. Decisions by algorithms are subject to\u0000different interpretations. Algorithms' behaviors are basis for the construal of\u0000moral assessment and standards. Yet we lack an understanding of how algorithms\u0000impact on social construction processes, and vice versa. Without such\u0000understanding, social construction processes may be disrupted and, eventually,\u0000may impede moral progress in society. We analyze the public discourse that\u0000emerged after a significant (five-fold) price-surge following the Brooklyn\u0000Subway Shooting on April 12, 2022, in New York City. There was much controversy\u0000around the two ride-hailing firms' algorithms' decisions. The discussions\u0000evolved around various notions of fairness and the algorithms' decisions'\u0000justifiability. Our results indicate that algorithms, even if not explicitly\u0000addressed in the discourse, strongly impact on constructing fairness\u0000assessments and notions. They initiate the exchange, form people's\u0000expectations, evoke people's solidarity with specific groups, and are a vehicle\u0000for moral crusading. However, they are also subject to adjustments based on\u0000social forces. We claim that the process of constructing notions of fairness is\u0000no longer just social; it has become a socio-algorithmic process. We propose a\u0000theory of socio-algorithmic construction as a mechanism for establishing\u0000notions of fairness and other ethical constructs.","PeriodicalId":501112,"journal":{"name":"arXiv - CS - Computers and Society","volume":"88 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141930916","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"AI Consciousness and Public Perceptions: Four Futures","authors":"Ines Fernandez, Nicoleta Kyosovska, Jay Luong, Gabriel Mukobi","doi":"arxiv-2408.04771","DOIUrl":"https://doi.org/arxiv-2408.04771","url":null,"abstract":"The discourse on risks from advanced AI systems (\"AIs\") typically focuses on\u0000misuse, accidents and loss of control, but the question of AIs' moral status\u0000could have negative impacts which are of comparable significance and could be\u0000realised within similar timeframes. Our paper evaluates these impacts by\u0000investigating (1) the factual question of whether future advanced AI systems\u0000will be conscious, together with (2) the epistemic question of whether future\u0000human society will broadly believe advanced AI systems to be conscious.\u0000Assuming binary responses to (1) and (2) gives rise to four possibilities: in\u0000the true positive scenario, society predominantly correctly believes that AIs\u0000are conscious; in the false positive scenario, that belief is incorrect; in the\u0000true negative scenario, society correctly believes that AIs are not conscious;\u0000and lastly, in the false negative scenario, society incorrectly believes that\u0000AIs are not conscious. The paper offers vivid vignettes of the different\u0000futures to ground the two-dimensional framework. Critically, we identify four\u0000major risks: AI suffering, human disempowerment, geopolitical instability, and\u0000human depravity. We evaluate each risk across the different scenarios and\u0000provide an overall qualitative risk assessment for each scenario. Our analysis\u0000suggests that the worst possibility is the wrong belief that AI is\u0000non-conscious, followed by the wrong belief that AI is conscious. The paper\u0000concludes with the main recommendations to avoid research aimed at\u0000intentionally creating conscious AI and instead focus efforts on reducing our\u0000current uncertainties on both the factual and epistemic questions on AI\u0000consciousness.","PeriodicalId":501112,"journal":{"name":"arXiv - CS - Computers and Society","volume":"25 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141930915","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Are Social Sentiments Inherent in LLMs? An Empirical Study on Extraction of Inter-demographic Sentiments","authors":"Kunitomo Tanaka, Ryohei Sasano, Koichi Takeda","doi":"arxiv-2408.04293","DOIUrl":"https://doi.org/arxiv-2408.04293","url":null,"abstract":"Large language models (LLMs) are supposed to acquire unconscious human\u0000knowledge and feelings, such as social common sense and biases, by training\u0000models from large amounts of text. However, it is not clear how much the\u0000sentiments of specific social groups can be captured in various LLMs. In this\u0000study, we focus on social groups defined in terms of nationality, religion, and\u0000race/ethnicity, and validate the extent to which sentiments between social\u0000groups can be captured in and extracted from LLMs. Specifically, we input\u0000questions regarding sentiments from one group to another into LLMs, apply\u0000sentiment analysis to the responses, and compare the results with social\u0000surveys. The validation results using five representative LLMs showed higher\u0000correlations with relatively small p-values for nationalities and religions,\u0000whose number of data points were relatively large. This result indicates that\u0000the LLM responses including the inter-group sentiments align well with actual\u0000social survey results.","PeriodicalId":501112,"journal":{"name":"arXiv - CS - Computers and Society","volume":"53 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141930921","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Moving beyond privacy and airspace safety: Guidelines for just drones in policing","authors":"Mateusz Dolata, Gerhard Schwabe","doi":"arxiv-2408.04684","DOIUrl":"https://doi.org/arxiv-2408.04684","url":null,"abstract":"The use of drones offers police forces potential gains in efficiency and\u0000safety. However, their use may also harm public perception of the police if\u0000drones are refused. Therefore, police forces should consider the perception of\u0000bystanders and broader society to maximize drones' potential. This article\u0000examines the concerns expressed by members of the public during a field trial\u0000involving 52 test participants. Analysis of the group interviews suggests that\u0000their worries go beyond airspace safety and privacy, broadly discussed in\u0000existing literature and regulations. The interpretation of the results\u0000indicates that the perceived justice of drone use is a significant factor in\u0000acceptance. Leveraging the concept of organizational justice and data\u0000collected, we propose a catalogue of guidelines for just operation of drones to\u0000supplement the existing policy. We present the organizational justice\u0000perspective as a framework to integrate the concerns of the public and\u0000bystanders into legal work. Finally, we discuss the relevance of justice for\u0000the legitimacy of the police's actions and provide implications for research\u0000and practice.","PeriodicalId":501112,"journal":{"name":"arXiv - CS - Computers and Society","volume":"194 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141930918","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Prioritising Response-able IP Practices in Digitization of Electoral Processes in Africa","authors":"Angella Ndaka, Samwel Oando, Eucabeth Majiwa","doi":"arxiv-2408.03690","DOIUrl":"https://doi.org/arxiv-2408.03690","url":null,"abstract":"Globally, people widely regard technology as a solution to global social\u0000problems. In a democratic society, its citizens view technology as a way to\u0000ensure commitment and sustaining the nation's democracy by allowing them to\u0000participate actively in the democratic process. However, despite the hype\u0000surrounding technology and development, many developing countries still\u0000experience democratic challenges. The democratic challenges have further led to\u0000barriers that shape the political landscape, resulting in delusion,\u0000disappointment, and failures in the democratic and public good processes, such\u0000as the electoral process. This paper explores the relationship between\u0000intellectual property (IP) practices and the adoption of digital technologies\u0000used in democratic electoral processes. Specifically, it examines how the\u0000prioritisation of IP by technology service providers can disrupt socio-material\u0000relationships in democratic electoral processes and outcomes. Because of the\u0000hard boundaries associated with IP it creates an environment where the systems\u0000are controlled solely by technology IP owners, while the consequences of\u0000electoral processes are borne by citizens. This questions the response-ability\u0000and trust-ability of digital technologies in running democratic processes.\u0000Drawing from the parallels in Kenya's general elections of 2017 and 2022, this\u0000paper illustrates how IP practices form a hard boundary that impels technology\u0000owners to micromanage electoral processes, leading to tensions that potentially\u0000create conflict. This finding can be used by decision-makers to adopt digital\u0000technologies and protect IP without compromising electoral processes and\u0000disrupting relationships in the wider society.","PeriodicalId":501112,"journal":{"name":"arXiv - CS - Computers and Society","volume":"84 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141930922","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}