On-line forums for the collaborative creation of bodies of information are a phenomenon of rising importance; the Wikipedia is one of the best-known examples. The open nature of such forums could benefit from a notion of reputation for its authors. Author reputation could be used to flag new contributions from low-reputation authors, and it could be used to allow only authors with good reputation to contribute to controversial or critical pages. A reputation system for the Wikipedia would also provide an incentive to give high-quality contributions. We present in this paper a novel type of contentdriven reputation system for Wikipedia authors. In our system, authors gain reputation when the edits and text additions they perform to Wikipedia articles are longlived, and they lose reputation when their changes are undone in short order. We have implemented the proposed system, and we have used it to analyze the entire Italian and French Wikipedias, consisting of a total of 691,551 pages and 5,587,523 revisions. Our results show that our notion of reputation has good predictive value: changes performed by low-reputation authors have a significantly larger than average probability of having poor quality, and of being undone.

Online communities, where users maintain lists of friends and express their preferences for items like movies, music, or books, are very popular. The web-based nature of this information makes it ideal for use in a variety of intelligent systems that can take advantage of the users ’ social and personal data. For those systems to be effective, however, it is important to understand the relationship between social and personal preferences. In this work we investigate features of profile similarity and how those relate to the way users determine trust. Through a controlled study, we isolate several profile features beyond overall similarity that affect how much subjects trust a hypothetical users. We then use data from FilmTrust, a real social network where users rate movies, and show that the profile features discovered in the experiment allow us to more accurately predict trust than when using only overall similarity. In this paper, we present these experimental results and discuss the potential implications for social networking and intelligent systems. 1.

In this paper we give an overview of the Foafing the Music system. The system uses the Friend of a Friend (FOAF) and Rich Site Summary (RSS) vocabularies for recommending music to a user, depending on her musical tastes. Music information (new album releases, related artists’ news and available audio) is gathered from thousands of RSS feeds —an XML format for syndicating Web content. On the other hand, FOAF documents are used to define user preferences. The presented system provides music discovery by means of: user profiling —defined in the user’s FOAF description—, context-based information —extracted from music related RSS feeds — and content-based descriptions —extracted from the audio itself. 1

Abstract. Social networks are a popular movement on the web. On the Semantic Web, it is simple to make trust annotations to social relationships. In this paper, we present a two level approach to integrating trust, provenance, and annotations in Semantic Web systems. We describe an algorithm for inferring trust relationships using provenance information and trust annotations in Semantic Web-based social networks. Then, we present an application, FilmTrust, that combines the computed trust values with the provenance of other annotations to personalize the website. The FilmTrust system uses trust to compute personalized recommended movie ratings and to order reviews. We believe that the results obtained with FilmTrust illustrate the success that can be achieved using this method of combining trust and provenance on the Semantic Web. 1

In many computing systems, information is produced and processed by many people. Knowing how much a user trusts a source can be very useful for aggregating, filtering, and ordering of information. Furthermore, if trust is used to support decision making, it is important to have an accurate estimate of trust when it is not directly available, as well as a measure of confidence in that estimate. This paper describes a new approach that gives an explicit probabilistic interpretation for confidence in social networks. We describe SUNNY, a new trust inference algorithm that uses a probabilistic sampling technique to estimate our confidence in the trust information from some designated sources. SUNNY computes an estimate of trust based on only those information sources with high confidence estimates. In our experiments, SUNNY produced more accurate trust estimates than the well known trust inference algorithm TIDALTRUST (Golbeck 2005), demonstrating its effectiveness.

Trust is a crucial basis for interactions among parties in large, open systems. Yet, the scale and dynamism of such systems make it infeasible for each party to have a direct basis for trusting another party. For this reason, the participants in an open system must share information about trust. However, they should not automatically trust such shared information. This paper studies the problem of propagating trust in multiagent systems. It describes a new algebraic approach, shows some theoretical properties of it, and empirically evaluates it on two social network datasets. This evaluation incorporates a new methodology that involves dealing with opinions in an evidential setting. 1

We investigate how social networks can be used in recommendation generation in taste related domains. Social Filtering (using social networks for neighborhood generation) is compared to Collaborative Filtering with respect to prediction accuracy in the domain of rating clubs. After reviewing background and related work, we present an extensive empirical study where over thousand participants from a social networking community where asked to provide ratings for clubs in Munich. We then compare a typical traditional CFapproach to a social recommender / social filtering approach where friends from the underlying social network are used as rating neighborhood and analyze the experiments statistically. Surprisingly, the social filtering approach outperforms the CF approach in all variants of the experiment. The implications of the experiment for professional and private-life collaborative environments and services where recommendations play a role are discussed. We conclude with future perspectives on social recommender systems, especially in upcoming mobile environments.

Social networks are a popular movement on the web. Trust can be used effectively on the Semantic Web as annotations to social relationships. In this paper, we present a two level approach to integrating trust, provenance, and annotations in Semantic Web systems. We describe an algorithm for inferring trust relationships using provenance information and trust annotations in Semantic Web-based social networks. Then, we present two applications that combine the computed trust values with the provenance of other annotations to personalize websites. The FilmTrust system uses trust to compute personalized recommended movie ratings and to order reviews. An open source intelligence portal, Profiles In Terror, also has a beta system that integrates social networks with trust annotations. We believe that these two systems illustrate a unique way of using trust annotations and provenance to process information on the Semantic Web. 1.

A drawback of traditional default logic is that there is no general mechanism for preferring one default rule over another. To remedy this problem, numerous default logics augmented with priority relations have been introduced. In this paper, we show how trust values, derived from web-based social networks, can be used to prioritize defaults. We provide a coupling between the method for computing trust values in social networks and the prioritized Reiter defaults of (Baader & Hollunder 1995), where specificity of terminological concepts is used to prioritize defaults. We compare our approach with specificity-based prioritization, and discuss how the two can be combined. Finally, we show how our approach can be applied to other variants of prioritized default logic.

To achieve better interoperability among intelligent applications, and to relieve knowledge engineers from the burden of developing ontologies from scratch, it is critical to reuse ontologies. However, there are two main reasons why the reuse of ontologies is rare: (1) current ontology repositories allow only simple keyword-based search facilities, and (2) even when a user finds an ontology, the information about the ontology quality and (re)usability is not available.