Mete Celik

Artificial bee colony data miner (ABC-Miner)

Data mining aims to discover interesting, non-trivial, and meaningful information from large datasets. One of the data mining tasks is classification, which aims to assign the given datasets to the most suitable classes. Classification rules are used in many domains such as medical sciences, banking, and meteorology. However, discovering classification rules is challenging due to large size and noisy structure of the datasets, and the difficulty of discovering general and meaningful rules. In the literature, there are several classical and heuristic algorithms proposed to mine classification rules out of large datasets. In this paper, a new and novel heuristic classification data mining approach based on artificial bee colony algorithm (ABC) was proposed (ABC-Miner). The proposed approach was compared with Particle Swarm Optimization (PSO) rule classification algorithm and C4.5 algorithm using benchmark datasets. The experimental results show the efficiency of the proposed method.

Anomaly detection in temperature data using DBSCAN algorithm

Anomaly detection is a problem of finding unexpected patterns in a dataset. Unexpected patterns can be defined as those that do not conform to the general behavior of the dataset. Anomaly detection is important for several application domains such as financial and communication services, public health, and climate studies. In this paper, we focus on discovery of anomalies in monthly temperature data using DBSCAN algorithm. DBSCAN algorithm is a density-based clustering algorithm that has the capability of discovering anomalous data. In the experimental evaluation, we compared the results of DBSCAN algorithm with the results of a statistical method. The analysis showed that DBSCAN has several advantages over the statistical approach on discovering anomalies.

Partial spatio-temporal co-occurrence pattern mining

Spatio-temporal co-occurrence patterns represent subsets of object-types that are often located together in space and time. The aim of the discovery of partial spatio-temporal co-occurrence patterns (PACOPs) is to find co-occurrences of the object-types that are partially present in the database. Discovering PACOPs is an important problem with many applications such as discovering interactions between animals in ecology, identifying tactics in battlefields and games, and identifying crime patterns in criminal databases. However, mining PACOPs is computationally very expensive because the interest measures are computationally complex, databases are larger due to the archival history, and the set of candidate patterns is exponential in the number of object-types. Previous studies on discovering spatio-temporal co-occurrence patterns do not take into account the presence period (i.e., lifetime) of the objects in the database. This paper defines the problem of mining PACOPs, proposes a new monotonic composite interest measure, and proposes novel PACOP mining algorithms. The experimental results show that the proposed algorithms are computationally more efficient than the naïve alternatives.

Discovering socially important locations of social media users

Abstract Socially important locations are places that are frequently visited by social media users in their social media life. Discovering socially interesting, popular or important locations from a location based social network has recently become important for recommender systems, targeted advertisement applications, and urban planning, etc. However, discovering socially important locations from a social network is challenging due to the data size and variety, spatial and temporal dimensions of the datasets, the need for developing computationally efficient approaches, and the difficulty of modeling human behavior. In the literature, several studies are conducted for discovering socially important locations. However, majority of these studies focused on discovering locations without considering historical data of social media users. They focused on analysis of data of social groups without considering each user’s preferences in these groups. In this study, we proposed a method and interest measures to discover socially important locations that consider historical user data and each user’s (individual’s) preferences. The proposed algorithm was compared with a naive alternative using real-life Twitter dataset. The results showed that the proposed algorithm outperforms the naive alternative.

CoABCMiner: An Algorithm for Cooperative Rule Classification System Based on Artificial Bee Colony

In data mining, classification rule learning extracts the knowledge in the representation of IF_THEN rule which is comprehensive and readable. It is a challenging problem due to the complexity of data sets. Various meta-heuristic machine learning algorithms are proposed for rule learning. Cooperative rule learning is the discovery process of all classification rules with a single run concurrently. In this paper, a novel cooperative rule learning algorithm, called CoABCMiner, based on Artificial Bee Colony is introduced. The proposed algorithm handles the training data set and discovers the classification model containing the rule list. Token competition, new updating strategy used in onlooker and employed phases, and new scout bee mechanism are proposed in CoABCMiner to achieve cooperative learning of different rules belonging to different classes. We compared the results of CoABCMiner with several state-of-the-art algorithms using 14 benchmark data sets. Non parametric statistical tests, such as Friedman test, post hoc test, and contrast estimation based on medians are performed. Nonparametric tests determine the similarity of control algorithm among other algorithms on multiple problems. Sensitivity analysis of CoABCMiner is conducted. It is concluded that CoABCMiner can be used to discover classification rules for the data sets used in experiments, efficiently.

Performance analysis of ABCMiner algorithm with different objective functions

Metaheuristic-based data mining algorithms are frequently used in literature for discovering meaningful rules out of huge datasets. However, in the design criteria of these algorithms, the choice of objective functions affects the performance of the algorithm and classification accuracy. ABCMiner is one of these algorithms and is a classification rule learning algorithm based on a swarm based metaheuristic algorithm, Artificial Bee Colony algorithm. In this paper, the performances of two different objective functions on ABCMiner are evaluated. The experimental evaluation is conducted using real datasets.

Mining periodic spatio-temporal co-occurrence patterns: A summary of results

Periodic spatio-temporal co-occurrence patterns (PECOPs) represent subsets of object-types that are often periodically located together in space and time. Discovering PECOPs is an important problem with many applications such as discovering interactions between animals and identifying tactics in games. However, mining PECOPs is computationally very expensive because the interest measures are computationally complex, databases are larger due to the archival history, and the set of candidate patterns is exponential in the number of object-types. In this paper, we define the problem of mining PECOPs, and propose a novel PECOP mining algorithm. The experimental results show that the proposed algorithm is computationally more efficient than the naïve alternatives.

Discovering socio-spatio-temporal important locations of social media users

Abstract Socio-spatio-temporal important locations (SSTILs) are places which are frequently visited by social media users in their social media history. Discovering SSTILs is important for several application domains, such as, recommender systems, advertisement applications, urban planning, etc. However, discovering SSTILs is challenging due to spatial, temporal, and social dimensions of the datasets, the lack of sufficient interest measures, and the need for developing computationally-efficient algorithms. In the literature, several methods are proposed to discover social important locations. However, these studies, usually, do not take into account temporal and social dimensions of the datasets and preferences of each user in a social group. In this study, we define SSTILs and SSTIL mining problem by taking into account spatial, temporal, and social dimensions of the social media datasets. We propose methods and interest measures to discover SSTILs efficiently based on both user and group preferences. The proposed algorithms were compared with a naive alternative using real-life Twitter dataset. The results showed that the proposed algorithms outperform the naive alternative.

Discovering socially similar users in social media datasets based on their socially important locations

Abstract Socially similar social media users can be defined as users whose frequently visited locations in their social media histories are similar. Discovering socially similar social media users is important for several applications, such as, community detection, friendship analysis, location recommendation, urban planning, and anomaly user and behavior detection. Discovering socially similar users is challenging due to dataset size and dimensions, spam behaviors of social media users, spatial and temporal aspects of social media datasets, and location sparseness in social media datasets. In the literature, several studies are conducted to discover similar social media users out of social media datasets using spatial and temporal information. However, most of these studies rely on trajectory pattern mining methods or take into account semantic information of social media datasets. Limited number of studies focus on discovering similar users based on their social media location histories. In this study, to discover socially similar users, frequently visited or socially important locations of social media users are taken into account instead of all locations that users visited. A new interest measure, which is based on Levenshtein distance, was proposed to quantify user similarity based on their socially important locations and two algorithms were developed using the proposed method and interest measure. The algorithms were experimentally evaluated on a real-life Twitter dataset. The results show that the proposed algorithms could successfully discover similar social media users based on their socially important locations.

Assessment of Multi Fragment Melting Analysis System (MFMAS) for the Identification of Food-Borne Yeasts

Multi Fragment Melting Analysis System (MFMAS) is a novel approach that was developed for the species-level identification of microorganisms. It is a software-assisted system that performs concurrent melting analysis of 8 different DNA fragments to obtain a fingerprint of each strain analyzed. The identification is performed according to the comparison of these fingerprints with the fingerprints of known yeast species recorded in a database to obtain the best possible match. In this study, applicability of the yeast version of the MFMAS (MFMAS-yeast) was evaluated for the identification of food-associated yeast species. For this purpose, in this study, a total of 145 yeast strains originated from foods and beverages and 19 standard yeast strains were tested. The DNAs isolated from these yeast strains were analyzed by the MFMAS, and their species were successfully identified with a similarity rate of 95% or higher. It was shown that the strains belonged to 43 different yeast species that are widely found in the foods. A clear discrimination was also observed in the phylogenetically related species. In conclusion, it might be suggested that the MFMAS-yeast seems to be a highly promising approach for a rapid, accurate, and one-step identification of the yeasts isolated from food products and/or their processing environments.