Tag Archives: Audun Jøsang

Trust and Reputation Management in Web-based Social Network

Author
Touhid Bhuiyan
Faculty
Science and Technology
Queensland University of Technology, Australia
Audun Jøsang
University of Oslo, Norway
Yue Xu
Faculty
Science and Technology
Queensland University of Technology, Australia

Trust and Reputation Management
in Web-based Social Network
For details please see the attached file:


Analysing Trust Transitivity and The Effects of Unknown Dependence

Author
Touhid Bhuiyan
Faculty
Science and Technology
Queensland University of Technology, Australia
Audun Jøsang
Faculty
Science and Technology
Queensland University of Technology, Australia
Abstract:
Trust can be used to improve online automated recommendation within a given domain. Trust transitivity is used to make it successful. But trust transitivity has different interpretations. Trust and trust transitivity; both are the human mental phenomenon and for this reason, there is no such thing as objective transitivity. Trust transitivity and trust fusion both are important elements in computational trust. This paper analyses the parameter dependence problem in trust transitivity and proposes some definitions considering the effects of base rate. In addition, it also proposes belief functions based on subjective logic to analyse trust transitivity of three specified cases with sensitive and insensitive based rate. Then it presents a quantitative analysis of the effects of unknown dependence problem in an interconnected network environment; such Internet.
Keywords: Trust management, Transitivity, Subjective logic, Base rate, Dependence.
For details please see the attached file:

A Review of Trust in Online Social Networks to Explore New Research Agenda

Author
Touhid Bhuiyan
Faculty
Science and Technology
Queensland University of Technology, Australia
Yue Xu
Faculty
Science and Technology
Queensland University of Technology, Australia
Audun Jøsang
Faculty
Science and Technology
Queensland University of Technology, Australia
Abstract – Trust has become important topic of research in many fields including sociology, psychology, philosophy, economics, business, law and of course in IT. In recent years, there is a dramatic growth in number and popularity of online social networks. There are many networks available with more than 100 million registered users such as Facebook, MySpace, QZone, Windows Live Spaces etc. People may connect,discover and share by using these  online social networks. The exponential growth of online communities in the area of social networks attracts the attention of the researchers about the importance of managing trust in online environment. The major challenge of the current online user is to identify the trustworthiness of the agents they are communicating. Trust can be calculated between two unknown agents who are not directly connected in a trust network rather connected somehow through the network. In this paper, we have reviewed the state-of-the-art research work on trust in online social network and discussed about the relevant research agenda.
Keywords: Online Social Networks, Trust, Reputation, Opinion, Recommender Systems.
For details please see the attached file:

The Potential Relationship between Trust and Interest Similarity

Author
Touhid Bhuiyan
Faculty
Science and Technology
Queensland University of Technology, Australia
Yue Xu
Science and Technology
Queensland University of Technology, Australia
Audun Jøsang
Faculty
Science and Technology
Queensland University of Technology, Australia
Abstract – Many online communities including the popular social networks offer automated recommendations to their users. This automated recommendations are normally generated using collaborative filtering systems based on the past ratings or opinions of the similar users. Alternatively, trust among the users in the network also can be used to find the neighbors while making recommendations. To obtain the optimum result, there must be a positive correlation exits between trust and interest similarity. In this paper we have presented the result of the survey on the relationship between trust and interest similarity. Our result supports the assumed hypothesis of positive relationship between the trust and interest similarity of the users.
Keywords: Trust, Interest, Similarity, Recommender System, Opinion, Social Networks.
For details please see the attached file:

An Analysis of Trust Transitivity Taking Base Rate into Account

Author
Touhid Bhuiyan
School of IT
Queensland University of Technology, Australia
Audun Jøsang
UNIK Graduate Center
University of Oslo, Norway
Yue Xu
School of IT
Queensland University of Technology, Australia
Abstract:
Trust transitivity, as trust itself, is a human mental phenomenon, so there is no such thing as objective transitivity, and trust transitivity therefore lends itself to different interpretations. Trust transitivity and trust fusion both are important elements in computational trust. This paper analyses the parameter dependence problem in trust transitivity and proposes some definitions considering the effects of base rate. In addition, it also proposes belief functions based on subjective logic to analyse trust transitivity of three specified cases with sensitive and insensitive based rate. Then it presents a quantitative analysis of the issue of exaggerated beliefs in Mass Hysteria based on subjective logic.
For details please see the attached file:

State-of-the-Art Review on Opinion Mining from Online Customers’ Feedback

Author
Touhid Bhuiyan
Queensland University of Technology, Australia
Yue Xu
Queensland University of Technology, Australia
Audun Jøsang
University of Oslo, Norway

Abstract:

Dealing with the ever-growing information overload in the Internet, Recommender Systemsare widely used online to suggest potential customers item they may like or find useful. Collaborative Filtering is the most popular techniques for Recommender Systems which collects opinions from customers in the form of ratings on items, services or service providers. In addition to the customer rating about a service provider, there is also a good number of online customer feedback information available over the Internet as customer reviews, comments, newsgroups post, discussion forums or blogs which is collectively called user generated contents. This information can be used to generate the public reputation of the service providers’. To do this, data mining techniques, specially recently emerged opinion mining could be a useful tool. In this paper we present a state of the art review of Opinion Mining from online customer feedback. We critically evaluate the existing work and expose cutting edge area of interest in opinion mining. We also classify the approaches taken by different researchers into several categories and sub-categories. Each of those steps is analyzed with their strength and limitations in this paper.

Key words: Data mining, Text mining,  pinion, Sentiment analysis, customer feedback

For details please see the attached file:

Integrating Trust with Public Reputation in Location-based Social Networks for Recommendation Making

Author
Touhid Bhuiyan
Faculty
Information Technology
Queensland University of Technology, Australia
Yue Xu
Faculty
Information Technology
Queensland University of Technology, Australia
Audun Jøsang
Faculty
Information Technology
Queensland University of Technology, Australia
Abstract:
 
The recent emergence of location-based social networking services is revolutionizing web-based social networking allowing users to share real-life experiences via geo-tagged user-generated multimedia content. One of the key challenges of the web-based social networks as an information sharing and exchanging channel is how to manage healthy relationships among community users and ensure the quality of the information shared and exchanged within the community, which holds a very significant importance to user satisfaction. Deciding whom and what information to trust is very difficult in environment where the users are unknown to each other. This paper investigates the possibilities of managing trust between the users of a web-based social network while recommending items to the members of the network. A novel framework is proposed to integrate trust among community members and public reputation of items to recommend the most appropriate items to a user of the network.
For details please see the attached file:

Combining Trust and Reputation Management for Web-Based Services

Author
Touhid Bhuiyan
Faculty
Information Technology
QUT, Brisbane, Australia
Audun Jøsang
University of Oslo, Norway
Yue Xu
University of Oslo, Norway
Abstract:
Services offered and provided through the Web have varying quality, and it is often difficult to assess the quality of a services before accessing and using it. Trust and reputation systems can be used in order to assist users in predicting and selecting the best quality services. This paper describes how Bayesian reputation systems can be combined with trustmodeling based on subjective logic to provide an integrated method for assessing the quality of online services. This will not only assist the user’s decision making, but will also provide an incentive for service providers to maintain high quality, and can be used as a sanctioning mechanism to discourage deceptive and low quality services.
For details please see the attached file:

Optimal Trust Network Analysis with Subjective Logic

Author
Touhid Bhuiyan
Faculty
Information Technology
QUT, Brisbane Australia
Audun Jøsang
University of Oslo Norway
Abstract:
Trust network analysis with subjective logic (TNA-SL) simplifies complex trust graphs into series-parallel graphs by removing the most uncertain paths to obtain a canonical graph. This simplification could in theory cause loss of information and thereby lead to sub-optimal results. This paper describes a new method for trust network analysis which is considered optimal because it does not require trust graph simplification, but instead uses edge splitting
to obtain a canonical graph. The new method is compared with TNA-SL, and our simulation shows that both methods produce equal results. This indicates that TNA-SL in fact also represents an optimal method for trust network analysis and that the trust graph simplification does not affect the result.
For details please see the attached file: