Browsing by Author "Zaamout, Khobaib"

Now showing 1 - 2 of 2
  • Thumbnail Image
    ItemOpen Access
    A Report on Propagative Influence: An Influence measure for directed and undirected networks
    (2018-03-15) Zaamout, Khobaib
    In this report, we present a new influence measure, namely Propagative Influence - PI. PI measure defines influence of a node in terms of its relationships. It suggests that a node's influence is a product of the interactions of nodes and is a state given to the node by the recognition and endorsements of others. We show that using PI will identify all influential nodes in a network in an effective and timely manner. The theory of this measure is very relevant and easily understood. We propose two variations of this measure, Static Propagative Influence - SPI and Dynamic Propagative Influence - DPI. Both variations use the same underlying algorithm but differ in influence propagation values. In this work, the abilities and effectiveness of our proposed measure is demonstrated through a series of examples on synthetic datasets which reflect real-world situations.
  • Thumbnail Image
    ItemOpen Access
    A Team Composition Approach For Social Crowdsourcing Communities
    (2020-09-22) Zaamout, Khobaib; Barker, Ken E.; Ruhe, Guenther; Tang, Anthony Hoi Tin
    This research takes place in an emerging paradigm of social computation that we name social crowdsourcing communities (SCCs). These are moderated online communities where members participate in collaborative activities (i.e. queries) designed to elicit their opinions concerning some topics, products, or services. This paradigm is distinct in that it combines the powers of crowdsourcing and social networking (SN) to allow for systematic querying of crowds and synthesizing response data (i.e. contributions) into coherent reports for decision-makers. SCCs consist of a beneficiary (i.e. the operators, the moderators, the analysts, and the organization that benefits from the reports), queries, a working crowd, and a platform where all activities occur. We show that it is possible to apply methods and techniques from existing fields to alleviate many of their challenges. One of these challenges is improving teamwork outcomes (i.e. contribution quality). Currently, SCC members, who are interested in a specific task, self-assemble into teams without considering any factors that may cause them to exhibit lower levels of productivity, participation, and contribution quality. The growth and query frequency restrictions imposed on these platforms by their operators to control operation costs further exacerbate this challenge. This thesis demonstrates how member behaviour can guide team formations and identifies specific behavioural characteristics related to improved team performance through an exploratory case study. It accomplishes this goal by capturing member behaviour in a model and using it to explore the compositions of existing teams. In doing so, this thesis identifies the specific compositions associated with increased team performance. The outcomes indicate the validity of this approach and provide a strong foundation for further investigation.