Athraa Jasim Mohammed

Weight-Based Firefly Algorithm for Document Clustering

Existing clustering techniques have many drawbacks and this includes being trapped in a local optima. In this paper, we introduce the utilization of a new meta-heuristics algorithm, namely the Firefly algorithm (FA) to increase solution diversity. FA is a nature-inspired algorithm that is used in many optimization problems. The FA is realized in document clustering by executing it on Reuters-21578 database. The algorithm identifies documents that has the highest light intensity in a search space and represents it as a centroid. This is followed by recognizing similar documents using the cosine similarity function. Documents that are similar to the centroid are located into one cluster and dissimilar in the other. Experiments performed on the chosen dataset produce high values of Purity and F-measure. Hence, suggesting that the proposed Firefly algorithm is a possible approach in document clustering.

Document Clustering Based on Firefly Algorithm

Document clustering is widely used in Information Retrieval however, existing clustering techniques suffer from local optima problem in determining the k number of clusters. Various efforts have been put to address such drawback and this includes the utilization of swarm-based algorithms such as particle swarm optimization and Ant Colony Optimization. This study explores the adaptation of another swarm algorithm which is the Firefly Algorithm (FA) in text clustering. We present two variants of FA; Weight- based Firefly Algorithm (WFA) and Weight-based Firefly Algorithm II (WFAII). The difference between the two algorithms is that the WFAII, includes a more restricted condition in determining members of a cluster. The proposed FA methods are later evaluated using the 20Newsgroups dataset. Experimental results on the quality of clustering between the two FA variants are presented and are later compared against the one produced by particle swarm optimization, K-means and the hybrid of FA and -K-means. The obtained results demonstrated that the WFAII outperformed the WFA, PSO, K-means and FA-Kmeans. This result indicates that a better clustering can be obtained once the exploitation of a search solution is improved.

A Newton’s Universal Gravitation Inspired Firefly Algorithm for Document Clustering

The divisive clustering has the advantage to build a hierarchical structure that is more efficient to represent documents in search engines. Its operation employs one of the partition clustering algorithms that leads to being trapped in a local optima. This paper proposes a Firefly algorithm that is based on Newton’s law of universal gravitation, known as Gravitation Firefly Algorithm (GFA), for document clustering. GFA is used to find centers of clusters based on objective function that maximizes the force between each document and an initial center. Upon identification of a center, the algorithm then locates documents that are similar to the center using cosine similarity function. The process of finding centers for new clusters continues by sorting the light intensity values of the balance documents. Experimental results on Reuters datasets showed that the proposed Newton inspired Firefly algorithm is suitable to be used for document clustering in text mining.

Determining Number of Clusters Using Firefly Algorithm with Cluster Merging for Text Clustering

Text mining, in particular the clustering is mostly used by search engines to increase the recall and precision of a search query. The content of online websites (text, blogs, chats, news, etc.) are dynamically updated, nevertheless relevant information on the changes made are not present. Such a scenario requires a dynamic text clustering method that operates without initial knowledge on a data collection. In this paper, a dynamic text clustering that utilizes Firefly algorithm is introduced. The proposed, aFAmerge, clustering algorithm automatically groups text documents into the appropriate number of clusters based on the behavior of firefly and cluster merging process. Experiments utilizing the proposed aFAmerge were conducted on two datasets; 20Newsgroups and Reuter’s news collection. Results indicate that the aFAmerge generates a more robust and compact clusters than the ones produced by Bisect K-means and practical General Stochastic Clustering Method (pGSCM).

Discovering optimal clusters using firefly algorithm

Existing conventional clustering techniques require a pre-determined number of clusters, unluckily; missing information about real world problem makes it a hard challenge. A new orientation in data clustering is to automatically cluster a given set of items by identifying the appropriate number of clusters and the optimal centre for each cluster. In this paper, we present the WFA_selection algorithm that originates from weight-based firefly algorithm. The newly proposed WFA_selection merges selected clusters in order to produce a better quality of clusters. Experiments utilising the WFA and WFA_selection algorithms were conducted on the 20Newsgroups and Reuters-21578 benchmark dataset and the output were compared against bisect K-means and general stochastic clustering method (GSCM). Results demonstrate that the WFA_selection generates a more robust and compact clusters as compared to the WFA, bisect K-means and GSCM.

Basic firefly algorithm for document clustering

The Document clustering plays significant role in Information Retrieval (IR) where it organizes documents prior to the retrieval process. To date, various clustering algorithms have been proposed and this includes the K-means and Particle Swarm Optimization. Even though these algorithms have been widely applied in many disciplines due to its simplicity, such an approach tends to be trapped in a local minimum during its search for an optimal solution. To address the shortcoming, this paper proposes a Basic Firefly (Basic FA) algorithm to cluster text documents. The algorithm employs the Average Distance to Document Centroid (ADDC) as the objective function of the search. Experiments utilizing the proposed algorithm were conducted on the 20Newsgroups benchmark dataset. Results demonstrate that the Basic FA generates a more robust and compact clusters than the ones produced by K-means and Particle Swarm Optimization (PSO).

Nature Inspired Data Mining Algorithm for Document Clustering in Information Retrieval

Document clustering is an important technique that has been widely employed in Information Retrieval (IR). Various clustering techniques have been reported, but the effectiveness of most techniques relies on the initial value of k clusters. Such an approach may not be suitable as we may not have prior knowledge on the collection of documents. To date, there are various swarm based clustering techniques proposed to address such problem, including this paper that explores the adaptation of Firefly Algorithm (FA) in document clustering. We extend the work on Gravitation Firefly Algorithm (GFA) by introducing a relocate mechanism that relocates assigned documents, if necessary. The newly proposed clustering algorithm, known as GFA_R, is then tested on a benchmark dataset obtained from the 20Newsgroups. Experimental results on external and relative quality metrics for the GFA_R is compared against the one obtained using the standard GFA and Bisect K-means. It is learned that by extending GFA to becoming GFA_R, a better quality clustering is obtained.

Experimental analysis of firefly algorithms for divisive clustering of web documents

This paper studies two clustering algorithms that are based on the Firefly Algorithm (FA) which is a recent swarm intelligence approach. We perform experiments utilizing the Newton’s Universal Gravitation Inspired Firefly Algorithm (GFA) and Weight-Based Firefly Algorithm (WFA) on the 20_newsgroups dataset. The analysis is undertaken on two parameters. The first is the alpha (α) value in the Firefly algorithms and latter is the threshold value required during clustering process. Results showed that a better performance is demonstrated by Weight-Based Firefly Algorithm compared to Newton’s Universal Gravitation Inspired Firefly Algorithm.

Blogs Search Engine Using RSS Syndication and Fuzzy Parameters

The rapid development of the internet eventually increases the number of internet users triggering the need for an intelligent search engine that is able to minimize the search on world wide web (WWW) and find relevant information as requested. To overcome the issue of finding relevant information as well as minimizing the search on WWW, this paper proposes a search engine that is specifically designed and built using RSS syndication and fuzzy Parameters to search for information contained in blogs. The blogs search engine consists of three main phases: 1) crawling using RSS feeds algorithm; 2) indexing weblogs algorithm; and 3) searching technique using fuzzy logic. In RSS crawling process, the RSS feeds need to be gathered to extract useful information such as title, links, time published, and description. Next, indexing weblogs uses the links to retrieve the blog sites for text processing and for constructing the indexing database. In order to retrieve such information requested or queried by any user, an interface is provided to enable the blog search based on keyword with associated degree of importance. The density of keyword is then computed from the indexing database. The rank of the pages is computed by using fuzzy weighted average. The experiment resulted in mean average precision of 81.7% of total system performance.