Students learning with intelligent tutoring systems (ITS) do not always trust system recommendations. One solution for this is explainable AI (XAI), which is shown to increase trust in AI. Our research focuses on how ...
详细信息
ISBN:
(纸本)9783031630279;9783031630286
Students learning with intelligent tutoring systems (ITS) do not always trust system recommendations. One solution for this is explainable AI (XAI), which is shown to increase trust in AI. Our research focuses on how students' personality traits affect their interactions with XAI, and how XAI affects students' trust and actions in an ITS. We evaluated this by adding XAI to SQL-Tutor and conducting a pilot study with 15 participants from an introductory database course. We found that personality traits affect students' interactions with XAI, and that students engaging with XAI trust the system more.
As users use an interactive application, they may, deliberately or inadvertently, use it in ways not anticipated by the developer. Unanticipated use of an application is considered an important and positive phenomenon...
详细信息
ISBN:
(纸本)9783031613616;9783031613623
As users use an interactive application, they may, deliberately or inadvertently, use it in ways not anticipated by the developer. Unanticipated use of an application is considered an important and positive phenomenon. It reveals clues to improve usability and enhance functionality, which were not conceived of in the initial design phase. However, developers often fail to benefit from such an opportunity due to the lack of sufficient knowledge about user behavior with respect to task completion. We propose an approach to discovering unexpected uses of an interactive application. In our approach, we choose aspect-oriented instrumentation to capture user interface events across architectural boundaries of an interactive application, perform model-based analysis to identify tasks and task sequences from preprocessed event traces, and conduct criteria-guided evaluation to discover unanticipated uses. Our goal is to provide tool support that is adaptable to an evolving application.
Layer 2 systems have received increasing attention due to their potential to scale the throughput of L1 blockchains. To avoid the cost of putting data on chain, these systems increasingly turn to off-chain data availa...
详细信息
ISBN:
(纸本)9783031477508;9783031477515
Layer 2 systems have received increasing attention due to their potential to scale the throughput of L1 blockchains. To avoid the cost of putting data on chain, these systems increasingly turn to off-chain data availability solutions such as data availability committees (DACs). However, placing trust on DACs conflicts with the goal of obtaining an L2 architecture whose security relies solely on the L1 chain. To eliminate such trust assumptions, we propose a DAC protocol that provides financial incentives to deter the DAC nodes from adversarial behavior such as withholding data upon request. We then analyze the interaction of rational DAC nodes and clients as a dynamic game, with a Byzantine adversary that can corrupt and bribe the participants. We also define a notion of optimality for the DAC protocols, inspired by fairness and economic feasibility. Our main result shows that our protocol is optimal and guarantees security with the highest possible probability under reasonable assumptions on the adversary.
This study investigated the impact of instructor gender (male versus female), type (human versus avatar), and lesson difficulty (easy and difficult) on the usability, trust, engagement, and comprehension of a virtual ...
详细信息
ISBN:
(纸本)9783031601248;9783031601255
This study investigated the impact of instructor gender (male versus female), type (human versus avatar), and lesson difficulty (easy and difficult) on the usability, trust, engagement, and comprehension of a virtual learning platform system. The study utilized animation software to create virtual male and female instructors and embedded them into a university psychology lecture on memory. Comprehension was evaluated through the use of multiple-choice tests, while usability, trust, and engagement were measured using the System Usability Scale, Trust in Automation Scale, and Paas Mental Effort Scale questionnaires, respectively. The results indicated no causal effect between instructor gender or type on comprehension, usability, and engagement. However, the study did reveal a significant effect of lesson difficulty on trust in the instructor, such that trust in the instructor was significantly higher following the difficult lesson. Implications of these findings pertain to the educational design of virtual classrooms.
Transaction fee markets are essential components of blockchain economies, as they resolve the inherent scarcity in the number of transactions that can be added to each block. In early blockchain protocols, this scarci...
详细信息
ISBN:
(纸本)9783031477508;9783031477515
Transaction fee markets are essential components of blockchain economies, as they resolve the inherent scarcity in the number of transactions that can be added to each block. In early blockchain protocols, this scarcity was resolved through a first-price auction in which users were forced to guess appropriate bids from recent blockchain data. Ethereum's EIP-1559 fee market reform streamlines this process through the use of a base fee that is increased (or decreased) whenever a block exceeds (or fails to meet) a specified target block size. Previous work has found that the EIP-1559 mechanism may lead to a base fee process that is inherently chaotic, in which case the base fee does not converge to a fixed point even under ideal conditions. However, the impact of this chaotic behavior on the fee market's main design goal - blocks whose long-term average size equals the target - has not previously been explored. As our main contribution, we derive near-optimal upper and lower bounds for the time-average block size in the EIP-1559 mechanism despite its possibly chaotic evolution. Our lower bound is equal to the target utilization level whereas our upper bound is approximate to 6% higher than optimal. Empirical evidence is shown in great agreement with these theoretical predictions. Specifically, the historical average was approximate to 2.9% larger than the target rage under Proof-of-Work and decreased to approximate to 2.0% after Ethereum's transition to Proof-of-Stake. We also find that an approximate version of EIP-1559 achieves optimality even in the absence of convergence.
This paper adds to the discussion of the nature of the scale on which even operates: on which dimension is such a scale based? The paper suggests a comparative analysis of two recent accounts, namely the degree-based ...
详细信息
ISBN:
(纸本)9783031608773;9783031608780
This paper adds to the discussion of the nature of the scale on which even operates: on which dimension is such a scale based? The paper suggests a comparative analysis of two recent accounts, namely the degree-based approach proposed by Y. Greenberg and the argumentative approach proposed by G. Winterstein. Following the former, even operates on a scale based on a gradable property. We claim that the argumentative approach of Winterstein may be interpreted as operating on the interval argumentative scale, which we suggest in this paper. The comparative analysis of the two approaches is conducted on three different levels: the level of data that they both account for (Sect. 3.2), on the structural level (Sect. 3.3), and on the functional level (3.4). At the end of the paper (cf. Sect. 4) we present some data that we believe to be challenging for both presented approaches;Sect. 5 draws conclusions.
Test evaluation plays vital roles in second language teaching, and hence requires more scientific and objective analysis methods. However, the domain of test analyses has not received enough attention, and this line r...
详细信息
ISBN:
(纸本)9789819742455;9789819742462
Test evaluation plays vital roles in second language teaching, and hence requires more scientific and objective analysis methods. However, the domain of test analyses has not received enough attention, and this line research is mainly based on surface data analyses with traditional and simple methods. This leaves plenty of room to exploring the underlying potentials of examinations. In general, test evaluation needs analyses on various aspects with more advanced models. Against this backdrop, the present case study tries to explore the test reliability and validity from comprehensive and objective perspectives by including different statistical analyses models. It focuses on analyzing the tests on the aspects of score distribution, correlation, reliability and validity, etc., especially with factor analysis and multitrait-multimethod matrix (MTMM). According to the results, the present test does not perform well in reliability and has average validity. The test assessment method could offer implications to improving English test designs.
In this paper we first investigated the relationship between game habits of primary school children and their school achievements and anxiety levels. Then, we investigated if children habitually playing video games at...
详细信息
ISBN:
(纸本)9783031490644;9783031490651
In this paper we first investigated the relationship between game habits of primary school children and their school achievements and anxiety levels. Then, we investigated if children habitually playing video games at home have an advantage when it comes to learning using educational games. In order to answer these questions, we exploited the data coming from the digital game-based learning (DGBL) intervention Happy Maths, a 6-week programme run in Irish primary schools aimed to increase maths abilities and decrease maths anxiety (MA). The dataset contained the academic achievements, the video game habits and the intervention data of 952 pupils. Our results show how playing games at home that are not age-appropriate was associated with higher MA and lower maths score, while time spent playing was associated with higher MA and lower literacy score. Regarding the efficacy of the DGBL intervention, there was no difference in the efficacy of the intervention between gamers and non-gamers. However, habitual video gamers were faster in executing their game moves, and they achieved higher scores, learning the game better. Overall, the study underlines the importance of playing age-appropriate games, and it provides evidence that, although kids playing video game might have a good advantage when it comes to educational games, the efficacy of such games is the same for gamers and non-gamers.
Communication involves transferring information from one place, person, or group to another, typically comprising a sender, a message, and a recipient. During interactions, humans exchange various social signals and i...
详细信息
ISBN:
(纸本)9783031601248;9783031601255
Communication involves transferring information from one place, person, or group to another, typically comprising a sender, a message, and a recipient. During interactions, humans exchange various social signals and information. There are two main types of communication: verbal and nonverbal. Even when verbal communication isn't feasible, nonverbal cues like gestures and body movements can facilitate communication. Hence, nonverbal communication plays a vital role in human-robot interaction. Implementing responsive strategies enables the creation of attentive listening systems. Recent research focuses on enhancing robot interactions through attentive listening systems, aiming to establish interfaces that foster affinity and a sense of connectionwith robots. The authors propose ARM-COMS for remote conferences, utilizing monitors as avatars for remote participants. This interface not only detects remote participants' head movements to coordinate actions but also responds to the voices of both remote participants and speakers, physically moving the monitor to enhance interaction. This paper explores these differences and discusses the philosophy behindARM-COMSfrom the perspective of attentive listening, aiming to enhance interaction with remote participants.
Expert human tutors can observe learner mistakes to understand their misconceptions and procedural errors. Highly capable, but opaque large language models have shown remarkable abilities across numerous domains, and ...
详细信息
ISBN:
(纸本)9783031606083;9783031606090
Expert human tutors can observe learner mistakes to understand their misconceptions and procedural errors. Highly capable, but opaque large language models have shown remarkable abilities across numerous domains, and may be useful for adaptive instruction in a variety of ways. Working with publicly available data from the National Assessment of Educational Progress, (388 questions selected from 4th, 8th and 12th grade math and science) we examined these three questions: 1) Do language models find the same problems difficult as students do? We found statistically significant, but small similarities in performance that differ somewhat by model. 2) Do language models have the same pattern of errors as students? Our findings reveal that, under the "minimal " prompts, the models often mirror students in choosing the same incorrect answers. However, this alignment decreases when prompt models used "chain of thoughts". 3) Can language models interpret and explain students' wrong answers? We presented frequently-chosen wrong answers toNAEP items to GPT-4 and an experienced science teacher, and compared their explanations. There was a good correspondence between these explanations, with 81% being fully or partially in agreement. Discussion focuses on how these capabilities can be used for test design and adaptive instruction.
暂无评论