Document clustering based on ant colony optimization algorithm has lately attracted the attention of many scholars throughout the globe. The aim of document clustering is to place similar content in one group, and non-similar contents in separate groups. In this article, by changing the behavior model of ant movement, we attempt to upgrade the standard ant's clustering algorithm. Ants' movement is completely random in the standard clustering algorithm. On the one hand, we improve the algorithm's efficiency by making ant movements purposeful, and on the other hand, by changing the rules of ant movement, we provide conditions so that the carrier ant moves to a location with intensive similarity with the carried component, and the noncarrier ant moves to a location where a component is surrounded by dissimilar components. We tested our proposed algorithm on a set of documents extracted from the 21578 Reuters Information Bank. Results show that the proposed algorithm on presents a better average performance compared to the standard ants clustering algorithm, and the K-means algorithm.
Abstractthe current article seeks to design and implement a new algorithm for data mining based on ant colony optimization algorithm, which is called Ant-Miner. Ant-Miner extracts classification rules from databases. In our article, we have presented a new version of Ant-Miner which is more efficient than its previous versions. The new version has been dubbed "Ant-Miner 4". We have modified the structure of the heuristic function used in Ant-Miner, implemented it based on the correction function of Laplace, and changed pheromone trail synchronization process in order to enable the redesigned system to produce rules with higher prediction power. In the proposed algorithm, we have tried to employ genetic algorithm to avoid local minimum points, produce a general optimized response, and determine the best values for the parameters. We tested Ant-Miner 4 and Ant-Miner 3 on four data sets, finding out that the new Ant-Miner has a better performance than the older version in terms of the accuracy of the extracted rules.
This paper proposes an algorithm for data mining called Ant-Miner (ant-colony-based data miner). The goal of Ant-Miner is to extract classification rules from data. The algorithm is inspired by both researches on the behavior of real ant colonies and some data mining concepts as well as principles. Recently research shows that ant colony optimization algorithm have been applied successfully to combinatorial optimization problems. In this paper we present an improvement to Ant-Miner. We compare the performance of new algorithm with before algorithm in two public domain data sets.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.