TruthTrust: Truth Inference-Based Trust Management Mechanism on a Crowdsourcing Platform

Zhou, Jian; Jin, Xing; Yu, Lanping; Li-Min, Xue; Ren, Yi

doi:10.3390/s21082578

Cited by 7 publications

(4 citation statements)

References 34 publications

(57 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Previous studies show that workers tend to "sabotage" customers with problematic behavior [25,26]. It is necessary to seek both worker protection and customer satisfaction by overcoming this vicious cycle through the mechanism of trust between customers and workers [27].…”

Section: Discussionmentioning

confidence: 99%

Analysis of Latent Classes and Factors Affecting them According to the Implementation of a Protection System Against Customer Bullying in Korea

Beak

Choi

Jung

2021

Preprint

View full text Add to dashboard Cite

Background: Instances of customers bullying employees at the workplace are increasing with the development of the service industry. Korea has established a worker protection system to prevent negative effects of customer bullying on workers’ health. This study identified the latent profile types of protection against customer bullying in workplaces, and determined their predictors.Methods: Data were collected over 28 days, from March 2 to 30, 2020. This study identified whether protection against customer bullying is implemented for workers in person-to-person services, the change effected by this protection, and worker monitoring scope. Data from 1,537 out of 1,550 participants were analyzed, excluding the missing values. Latent classes were identified using Mplus 8.5 for data analysis, and the multinomial logistic regression analysis was performed to review the predictors.Results: The results are as follows. There were four types of latent classes: the lagging type, the medium type, the preventive type, and the excellent type. Variables predicting these types included age, gender, service period, position, occupational category, worker monitoring scope, decrease in the number of customers causing problems, decrease in disputes with customers, and worker satisfaction. Conclusions: First, a system to prevent customer bullying must be adopted in the service industry. Second, there must be a way to improve trust between customers and workers when protective measures are being adopted. Third, managers must establish a system that can both protect workers against bullying and provide customer satisfaction.

show abstract

Section: Discussionmentioning

confidence: 99%

Analysis of Latent Classes and Factors Affecting them According to the Implementation of a Protection System Against Customer Bullying in Korea

Beak

Choi

Jung

2021

Preprint

View full text Add to dashboard Cite

show abstract

“…However, it is very prone to adversarial attacks and does not ensure the quality of crowdsourced answers. There are several methods for inferring the truth of tasks and for performing aggregation of submissions, such as [9,[15][16][17][18]. They are classified based on the underlying techniques as direct computation [15], probability-based methods [19], optimization methods [20], and neural network-based (NN) methods [21].…”

Section: Truth Inference and Quality Control In Crowdsourcingmentioning

confidence: 99%

Classifying Workers for Mitigating Adversarial Attacks in Crowdsourcing

First,

Second

2023

Preprint

View full text Add to dashboard Cite

Crowdsourcing is adopted as a fast and cost-effective system for human computation and acquiring data for training models in machine learning. Although crowdsourcing has broad applicability, it still has the following challenges. Normally, workers have access to crowdsourcing platforms with simple authentication mechanisms. As a result, malicious workers may get into the system and submit unreliable answers rendering the platform fraudulent. Moreover, when workers perform tasks, their lack of expertise and the difficulty of tasks affects the accuracy. Truth inference benefits the data quality and worker reliability. A truth inference algorithm estimates workers' trustworthiness and correctness of the answers from their responses. Besides, it helps in filtering out low-quality answers. However, due to adversarial attacks by malicious workers, ground truth inference algorithms perform poorly. This research considers how to defend against adversarial attacks on crowdsourcing platforms and improve the truth inference process. The proposed method estimates the trust and reliability scores of the workers and classifies them as normal and malicious workers. Based on this classification , tasks are assigned to the workers. Moreover, these predicted scores are used for inferring the correct answers, thereby improving the ground truth inference. Experiments consistently show that the proposed truth inference method is tolerant to adversarial attacks with competent accuracy.

show abstract

“…An approach is proposed in [38] to evaluate the trustworthiness of workers in crowdsourcing platforms such as Amazon Mechanical Turk. The approach specifically targets collusion attacks in crowdsourcing platforms.…”

Section: Trust Assessment Based On Previous Experiencesmentioning

confidence: 99%

Multi-Use Trust in Crowdsourced IoT Services

Bahutair¹,

Bouguettaya²,

Neiat³

2022

Preprint

View full text Add to dashboard Cite

We introduce the concept of adaptive trust in crowdsourced IoT services. It is a customized fine-grained trust tailored for specific IoT consumers. Usage patterns of IoT consumers are exploited to provide an accurate trust value for service providers. A novel adaptive trust management framework is proposed to assess the dynamic trust of IoT services. The framework leverages a novel detection algorithm to obtain trust indicators that are likely to influence the trust level of a specific IoT service type. Detected trust indicators are then used to build service-to-indicator model to evaluate a service's trust at each indicator. Similarly, a usage-to-indicator model is built to obtain the importance of each trust indicator for a particular usage scenario. The per-indicator trust and the importance of each trust indicator are utilized to obtain an overall value of a given service for a specific consumer. We conduct a set of experiments on a real dataset to show the effectiveness of the proposed framework.

show abstract

TruthTrust: Truth Inference-Based Trust Management Mechanism on a Crowdsourcing Platform

Cited by 7 publications

References 34 publications

Analysis of Latent Classes and Factors Affecting them According to the Implementation of a Protection System Against Customer Bullying in Korea

Analysis of Latent Classes and Factors Affecting them According to the Implementation of a Protection System Against Customer Bullying in Korea

Classifying Workers for Mitigating Adversarial Attacks in Crowdsourcing

Multi-Use Trust in Crowdsourced IoT Services

Contact Info

Product

Resources

About