The k-clique problem is identifying the largest complete subgraph of size k on a network, and it has many applications in Social Network Analysis (SNA), coding theory, geometry, etc. Due to the NP-Complete nature of the problem, the meta-heuristic approaches have raised the interest of the researchers and some algorithms are developed. In this paper, a new algorithm based on the Bat optimization approach is developed for finding the maximum k-clique on a social network to increase the convergence speed and evaluation criteria such as Precision, Recall, and F1-score. The proposed algorithm is simulated in Matlab® software over Dolphin social network and DIMACS dataset for k = 3, 4, 5. The computational results show that the convergence speed on the former dataset is increased in comparison with the Genetic Algorithm (GA) and Ant Colony Optimization (ACO) approaches. Besides, the evaluation criteria are also modified on the latter dataset and the F1-score is obtained as 100% for k = 5.

中文翻译：

---

使用蝙蝠优化算法发现社交网络上的最大k值

k clique问题正在识别网络上大小为k的最大的完整子图，它在社交网络分析（SNA），编码理论，几何等方面具有许多应用。由于问题的NP完全性质，元启发式方法引起了研究人员的兴趣，并开发了一些算法。在本文中，开发了一种基于蝙蝠优化方法的新算法，用于查找社交网络上的最大k值，以提高收敛速度和评估标准，例如Precision，Recall和F1-score。该算法在Matlab®软件中通过Dolphin社交网络和DIMACS数据集进行了仿真，其中k = 3、4、5。计算结果表明，与遗传算法和蚁群优化方法相比，前一数据集的收敛速度有所提高。此外，评估标准也在后一个数据集上进行了修改，对于k = 5，F1得分为100％。