Downloads: 174 | Views: 218

Research Paper | Management | Canada | Volume 8 Issue 4, April 2019

Presenting a New Model for the Management and Selection of Experts in Big Data

Nasrin Hashempour | Parisa Safarzadeh | Hanieh Raoofifard

Abstract: Finding a group of experts is one of the defined issues in the social fields. The aim is to appoint a group of experts, each having specific skills to do a variety of tasks, which for any of these tasks they require some particular expertise. This appointment is considered as NP-Hard issue, and its exact solution is exponential. Most-used cases for the selection of experts in Big Data apply the top-down greedy algorithm. The problem in these methods is the lack of attention to the specific skills and increase in run time when using the greedy algorithm; because sorting is based on more skills of each reviewer and there is no consideration to the number of skills from a special type. In this paper, we have used a model based on knapsack problem theory as well as applying an artificial bee colony algorithm to appoint a group of experts to review scientific papers. In this study, the algorithm continues until the selection of a set of the experts who cover all the skills required to reviewing papers; therefore, scoring the reviewers at each stage plays a significant role in determining the near-optimal set of answers. Also, to score the experts in Big Data, we have considered the two criteria of coverage and specific skills of experts. In the assessment section, the result of the implementation of the proposed algorithm on a standard data set has been investigated, and the results were compared with the greedy algorithm. The assessment results have shown that the proposed algorithm in terms of time and accuracy of selected experts, was superior to the greedy algorithm.

Keywords: Big Data, Experts, Reviewing, Scientific papers, Knapsack problem, Artificial bee colony algorithm, Coverage, Being special

Edition: Volume 8 Issue 4, April 2019,

Pages: 395 - 402