Artificial Bee Colony (ABC) is a recent meta-heuristic approach. In this paper we face the problem of clustering by ABC and we model a further bee role in the colony, performed by inspector bee. This model conforms with real honey bee colony, indeed, in nature some bees among the foraging ones are called inspectors because they preserve the colony’s history and historical information related to food sources. We experiment inspector behavior in ABC and compare the solution to traditional clustering algorithm. Finally, the effect of colony size is investigated and experimental results are discussed.

This paper proposes a new formulation of Artificial Bee Colony (ABC) in order to address clustering problems. The proposed algorithm models the inspector bee within the colony. It is tested for some benchmarks and is adopted to a real-world problem in Transportation System domain. In particular, we propose a clustering problem for the identification of vehicle usage in Poste Italiane by grouping together those vehicles with same features as fuel economies, frequency and value of refueling activities.

Several approaches for recommending products to the users are proposed in literature, and collaborative filtering has been proved to be one of the most successful techniques. Some issues related to the quality of recommendation and to computational aspects still arise (e.g., cold-start recommendations). In this paper, we investigate the application of model-based Collaborative Filtering (CF) techniques and in particular propose a clustering CF framework and two clustering CF algorithms: Item-based Fuzzy Clustering Collaborative Filtering (IFCCF) and Trust-aware Clustering Collaborative Filtering (TRACCF). We compare several approaches by means of Epinions, MovieLens, Jester, and Poste Italiane datasets (with real customers). Experimental results show an increased value of coverage of the recommendations provided by TRACCF without affecting recommendation quality. Moreover, trust information guarantees high level recommendation for different users.

Identifying the products to suggest to the target customer in order to best fit his profile of interests is a challenging task in the Social Commerce domain. In this paper we investigate the application of different model-based Collaborative Filtering (CF) techniques and we propose a framework able to provide different suggestions to the customers. The framework is aimed at expanding the customer suggestions in order to guarantee a serendipitous discovery of wished products and at the same time, it is aimed at assisting merchants in discovering potential interests of their customers. We compare several approaches by means of Epinions, MovieLens and Poste Italiane dataset (with real customers). Experimental results show an increased value of coverage of the recommendations provided by our approach without affecting recommendation quality. Moreover, personalized strategy guarantees a high level recommendation for different user navigation behaviors.

Identifying a customer profile of interest is a challenging task for sellers. Preferences and profile features can range during the time in accordance with current trends. In this paper we investigate the application of different model-based Collaborative Filtering (CF) techniques and in particular propose a trusted approach to user-based clustering CF. We propose a Trust-aware Clustering Collaborative Filtering and we compare several approaches by means of Epinions, which contains explicit trust statements, and MovieLens dataset, where we have implicitly defined a trust information. Experimental results show an increased value of coverage of the recommendations provided by our approach without affecting recommendation quality. To conclude, we introduce a tool, based on recommender systems, able to assist merchants in delivering special offers or in discovering potential interests of their customers. This tool allows each merchant to identify the products to suggest to the target customer in order to best fit his profile of interests.

In this paper, we investigated the application of trust in an e-Commerce system. In a B2C scenario of an e-Commerce system, a game model is proposed in order to investigate the best strategies for merchants and customers. Preliminary investigation outlines the benefits of trust information in the proposed game and preliminary results, conducted on a transactional database, shows an increased value of sensitivity of provided recommendations to the customers, entailing an higher customer loyalty. Future works are aimed at validating this findings by means of a larger real dataset.

Predicting user preferences is a challenging task. Different approaches for recommending products to the users are proposed in literature and collaborative filtering has been proved to be one of the most successful techniques. Some issues related to the quality of recommendation and to computational aspects still arise (e.g., scalability and cold-start recommendations). In this paper, we propose an Item-based Fuzzy Clustering Collaborative Filtering (IFCCF) in order to ensure the benefits of a model-based technique improving the quality of suggestions. Experimentation led by predicting ratings of MovieLens and Jester users makes this promising and worth to be further investigated in a cross-domain dataset.

Recommendation systems are commonly used for suggesting products or services. Among different existing techniques, Model-Based Collaborative Filtering (MBCF) approaches have been proven to address scalability and cold-starting problems that often arise. In this paper we investigate two MBCF algorithms: Self-Organizing Maps (SOM) for Collaborative Filtering and Item-based Fuzzy Clustering Collaborative Filtering (IFCCF). These two techniques have been selected because preliminary results have proven that when applied to the clustering of users or items the quality of the recommendation system increases with respect to the k-means. Within recommendation systems, no comparison of these two techniques exists. Therefore, our experimentation is aimed at comparing these two techniques by means of MovieLens and Jester dataset in order to provide a guideline for their implementation in the e-Commerce domain.