Seyed-Mehdi-Reza Beheshti

Provenance refers to the documentation of an object's lifecycle. This documentation (often represented as a graph) should include all the information necessary to reproduce a certain piece of data or the process that led to it. In a dynamic world, as data changes, it is important to be able to get a piece of data as it was, and its provenance graph, at a certain point in time. Supporting time-aware provenance querying is challenging and requires: (i) explicitly representing the time information in the provenance graphs, and (ii) providing abstractions and efficient mechanisms for time-aware querying of provenance graphs over an ever growing volume of data. The existing provenance models treat time as a second class citizen (i.e. as an optional annotation). This makes time-aware querying of provenance data inefficient and sometimes inaccessible. We introduce an extended provenance graph model to explicitly represent time as an additional dimension of provenance data. We also provide a query language, novel abstractions and efficient mechanisms to query and analyze timed provenance graphs. The main contributions of the paper include: (i) proposing a Temporal Provenance Model (TPM) as a timed provenance model; and (ii) introducing two concepts of timed folder, as a container of related set of objects and their provenance relationship over time, and timed paths, to represent the evolution of objects tracing information over time, for analyzing and querying TPM graphs. We have implemented the approach on top of FPSPARQL, a query engine for large graphs, and have evaluated for querying TPM models. The evaluation shows the viability and efficiency of our approach.

The ability to efficiently find relevant subgraphs and paths in a large graph to a given query is important in many applications including scientific data analysis, social networks, and business intelligence. Currently, there is little support and no efficient approaches for expressing and executing such queries. This paper proposes a data model and a query language to address this problem. The contributions include supporting the construction and selection of: (i) folder nodes, representing a set of related entities, and (ii) path nodes, representing a set of paths in which a path is the transitive relationship of two or more entities in the graph. Folders and paths can be stored and used for future queries. We introduce FPSPARQL which is an extension of the SPARQL supporting folder and path nodes. We have implemented a query engine that supports FPSPARQL and the evaluation results shows its viability and efficiency for querying large graph datasets.

Worker selection is a significant and challenging issue in crowdsourcing systems. Such selection is usually based on an assessment of the reputation of the individual workers participating in such systems. However, assessing the credibility and adequacy of such calculated reputation is a real challenge. In this paper, we propose an analytic model which leverages the values of the tasks completed, the credibility of the evaluators of the results of the tasks and time of evaluation of the results of these tasks in order to calculate an accurate and credible reputation rank of participating workers and fairness rank for evaluators. The model has been implemented and experimentally validated.

Online rating systems are subject to malicious behaviors mainly by posting unfair rating scores. Users may try to individually or collaboratively promote or demote a product. Collaborating unfair rating 'collusion' is more damaging than individual unfair rating. Although collusion detection in general has been widely studied, identifying collusion groups in online rating systems is less studied and needs more investigation. In this paper, we study impact of collusion in online rating systems and asses their susceptibility to collusion attacks. The proposed model uses a frequent itemset mining algorithm to detect candidate collusion groups. Then, several indicators are used for identifying collusion groups and for estimating how damaging such colluding groups might be. Also, we propose an algorithm for finding possible collusive subgroup inside larger groups which are not identified as collusive. The model has been implemented and we present results of experimental evaluation of our methodology.