A NEW APPROACH TO DEVELOPMENT OF NSGAII AND NRGA FOR SOLVING COMMUNITY DETECTION PROBLEM
Message:
Article Type:
Research/Original Article (دارای رتبه معتبر)
Abstract:

Detecting community structures is applicable in a wide range of scienti c elds such as biological and social sciences. Communitydetectionisoneofthemostrenowned problems in the eld of social networks mining. Thus, many methods have been introduced and developed in order to meet diverse needs of community detection. The aim of community detection is to partition the network in such a way that relations between components of network are dense. Since the relations between the members of partitions are strong, it is possible to consider them as a community or a cluster. In this paper, we have considered community detection as a multiobjective problem. The objective functions are modularity and community scores, which are two of the most well-known objectives in the literature. In order to optimize these objective functions, two algorithms, which are the enhanced versions of NSGAII and NRGA, have been proposed. These methods use a greedy algorithm to obtain initial population. Moreover, new crossover and mutation operators have been designed. The crossover operator is based on closeness of nodes. The mutation operator is based on TOPSIS method. The proposed crossover and mutation operators always generate feasible solutions. Furthermore, the closeness index helps to form distinct and high quality communities. We have compared the performance of the proposed methods with those of classical NSGAII, NRGA, and a well-known method called MOGA-Net by conducting several numerical experiments in six real-world networks. The experiments split into two parts. In the rst part, we have compared the solutions of these ve algorithms regarding the values of objective functions. The second part is dedicated to the comparisons made based on various multi-objective metrics. We have considered spacing, generational distance, inverted generational distance, set coverage, normalized mutual information, computation time, and the number of nondominated solutions obtained by each method. In order to ensure that the solutions obtained by the proposed algorithms are signi cantly better than the ones provided by the other three methods, we have conducted several two-sample t-tests. The results showed signi cant improvement, and the proposed algorithms outperformed the other three methods regarding various criteria.

Language:
Persian
Published:
Industrial Engineering & Management Sharif, Volume:34 Issue: 1, 2018
Pages:
101 to 115
magiran.com/p1894385  
دانلود و مطالعه متن این مقاله با یکی از روشهای زیر امکان پذیر است:
اشتراک شخصی
با عضویت و پرداخت آنلاین حق اشتراک یک‌ساله به مبلغ 990,000ريال می‌توانید 70 عنوان مطلب دانلود کنید!
اشتراک سازمانی
به کتابخانه دانشگاه یا محل کار خود پیشنهاد کنید تا اشتراک سازمانی این پایگاه را برای دسترسی نامحدود همه کاربران به متن مطالب تهیه نمایند!
توجه!
  • حق عضویت دریافتی صرف حمایت از نشریات عضو و نگهداری، تکمیل و توسعه مگیران می‌شود.
  • پرداخت حق اشتراک و دانلود مقالات اجازه بازنشر آن در سایر رسانه‌های چاپی و دیجیتال را به کاربر نمی‌دهد.
دسترسی سراسری کاربران دانشگاه پیام نور!
اعضای هیئت علمی و دانشجویان دانشگاه پیام نور در سراسر کشور، در صورت ثبت نام با ایمیل دانشگاهی، تا پایان فروردین ماه 1403 به مقالات سایت دسترسی خواهند داشت!
In order to view content subscription is required

Personal subscription
Subscribe magiran.com for 50 € euros via PayPal and download 70 articles during a year.
Organization subscription
Please contact us to subscribe your university or library for unlimited access!