Web Usage Mining Approaches for Web Page Recommendation

The technology behind personalization or Web page recommendation has undergone tremendous changes, and several Web-based personalization systems have been proposed in recent years. The main goal of Web personalization is to dynamically recommend Web pages based on online behavior of users. Although personalization can be accomplished in numerous ways, most Web personalization techniques fall into four major categories: decision rule-based filtering, content-based filtering, and collaborative filtering and Web usage mining. Decision rule-based filtering reviews users to obtain user demographics or static profiles, and then lets Web sites manually specify rules based on them. It delivers the appropriate content to a particular user based on the rules. However, it is not particularly useful because it depends on users knowing in advance the content that interests them. Content-based filtering relies on items being similar to what a user has liked previously. Collaborative filtering, also called social or group filtering, is the most successful personalization technology to date. Most successful recommender systems on the Web typically use explicit user ratings of products or preferences to sort user profile information into peer groups. It then tells users what products they might want to buy by combining their personal preferences with those of like-minded individuals. However, collaborative filtering has limited use for a new product that no one has seen or rated, and content-based filtering to obtain user profiles might miss novel or surprising information. Additionally, traditional Web personalization techniques, including collaborative or content-based filtering, have other problems, such as reliance on subject user ratings and static profiles or the inability to capture richer semantic relationships among Web objects. To overcome these shortcomings, the new Web personalization tool, nonintrusive personalization, attempts to increasingly incorporate Web usage mining techniques. Web usage mining can help improve the scalability, accuracy, and flexibility of recommender systems. Thus, Web usage mining can reduce the need for obtaining subjective user ratings or registration-based personal preferences. This chapter provides a survey of Web usage mining approaches.

Hybrid Neural Genetic Architecture

The objective is a neural-based feature selection in intelligent recommender systems. In particular, a hybrid neural genetic architecture is modeled based on human nature, interactions, and behaviour. The main contribution of this chapter is the development of a novel genetic algorithm based on human nature, interactions, and behaviour. The novel genetic algorithm termed “Buabin Algorithm” is fully integrated with a hybrid neural classifier to form a Hybrid Neural Genetic Architecture. The research presents GA in a more attractive manner and opens up the various departments of a GA for active research. Although no scientific experiment is conducted to compare network performance with standard approaches, engaged techniques reveal drastic reductions in genetic operator operations. For illustration purposes, the UCI Molecular Biology (Splice Junction) dataset is used. Overall, “Buabin Algorithm” seeks to integrate human related interactions into genetic algorithms as imitate human genetics in recommender systems design and understand underlying datasets explicitly.

Hybrid Neural Architecture for Intelligent Recommender System Classification Unit Design

The objective is intelligent recommender system classification unit design using hybrid neural techniques. In particular, a neuroscience-based hybrid neural by Buabin (2011a) is introduced, explained, and examined for its potential in real world text document classification on the modapte version of the Reuters news text corpus. The so described neuroscience model (termed Hy-RNC) is fully integrated with a novel boosting algorithm to augment text document classification purposes. Hy-RNC outperforms existing works and opens up an entirely new research field in the area of machine learning. The main contribution of this book chapter is the provision of a step-by-step approach to modeling the hybrid system using underlying concepts such as boosting algorithms, recurrent neural networks, and hybrid neural systems. Results attained in the experiments show impressive performance by the hybrid neural classifier even with a minimal number of neurons in constituting structures.

Intelligent Techniques in Recommender Systems and Contextual Advertising

Information Filtering deals with the problem of selecting relevant information for a given user, according to her/his preferences and interests. In this chapter, the authors consider two ways of performing information filtering: recommendation and contextual advertising. In particular, they study and analyze them according to a unified view. In fact, the task of suggesting an advertisement to a Web page can be viewed as the task of recommending an item (the advertisement) to a user (the Web page), and vice versa. Starting from this insight, the authors propose a content-based recommender system based on a generic solution for contextual advertising and a hybrid contextual advertising system based on a generic hybrid recommender system. Relevant case studies have been considered (i.e., a photo recommender and a Web advertiser) with the goal of highlighting how the proposed approach works in practice. In both cases, results confirm the effectiveness of the proposed solutions.

Building Recommender Systems for Network Intrusion Detection Using Intelligent Decision Technologies

This chapter presents an overview of the field of recommender systems and describes the current generation of recommendation methods with their limitations and possible extensions that can improve the capabilities of the recommendations made suitable for a wide range of applications. In recent years, machine learning algorithms have been considered to be an important part of the recommendation process to take intelligent decisions. The chapter will explore the application of such techniques in the field of network intrusion detection in order to examine the vulnerabilities of different recommendation techniques. Finally, the authors outline some of the major issues in building secure recommendation systems in identifying possible network intrusions.

A Recommender System Supporting Teachers to Author Learning Sessions in Decision-Making

This chapter pays attention to the automatic generation and recommendation of teaching materials for teachers who do not have enough time to learn how to use authoring tools for the creation of materials to support their courses. To overcome the difficulties, the research is intended to solve the problem of time needed to create adapted case studies for teaching decision-making in network design. Another goal is to reduce the time required to learn the use of an authoring tool to create teaching materials. Thus, the author presents an assistant that provides adapted help for teachers, generates examples automatically, verifies that any generated example fits in the class of examples used by the teacher, and recommends personalized examples according to each teacher’s preferences. He studies the use of data related to teachers to support the recommendation of teaching materials and the adaptation of Web-based support. The automatic generation and test of examples of network topologies are based on a probabilistic model, and the recommendation is based on Bayesian classification. This investigation also looks at problems related to the application of Artificial Intelligence (AI) to support teachers in authoring learning sessions for Adaptive Educational Hypermedia (AEH).

Building Efficient Assessment Applications with Personalized Feedback

The aim of this chapter is to provide a model for requirement specification, useful in developing efficient e-assessment applications with personalized feedback, which is enhanced by calling a recommender engine. The research was done in the context of using educational technology to facilitate learning processes. The data used to build the requirement model was collected from a set of interviews with the users and creators of an e-assessment application in project management. Requirement analysis assumes human effort and thus introduces uncertainties. To minimize the subjective factor, the data extracted from interviews with the users and the developers of the existing e-assessment application are clustered using a fuzzy logic solution into classes of requirements. These classes are the units of the model. The connections between classes are also mentioned: relations such as “if-then,” “switch,” or” contains” are explained. The requirements analysis conducts a smart set of specifications, obtained in a collaborative manner, useful for the design of e-assessment applications in project management or other similar domains.

Newly Developed Nature-Inspired Algorithms and their Applications to Recommendation Systems

Because of their mathematical backgrounds and coherent structures, Artificial Intelligent-based methods and techniques are often used to find solutions for different types of problems encountered. In the related context, nature-inspired algorithms are also important for providing more accurate solutions. Because of their nature-based, flexible process structures, the related algorithms can be applied to different types of problems. At this point, recommendation systems are one of the related problem areas where nature-inspired algorithms can be used to get better results. In the literature there are many research studies that are based on using nature-inspired algorithms within recommendation systems. This chapter aims to discuss usage of some newly developed nature-inspired algorithms in typical recommendation systems. In this aim, features and functions of some new nature-inspired algorithms will be explained first. Later, using the related algorithms in recommendation systems will be discussed. Following that, there will be a discussion on future of nature-inspired algorithms and also their roles in the recommendation approach or system-based applications.

A Fuzzy Clustering Approach for Segmenting Retail Industry

The foremost theme of this chapter is to utilize the subtractive clustering concept for defining the market boundaries in the fuzzy-based segmentation. In this sense, the present work starts by analyzing the importance of segmenting the shoppers on the basis of store image. After reviewing the segmentation literature, the authors performed a segmentation analysis of retail shoppers in India. Researchers often use clustering analysis as a tool in market segmentation studies, the results of which often end with a crisp partitioning form, where one member cannot belong to two or more groups. This indicates that different segments overlap with one another. This chapter integrates the concept of application of subtractive clustering in fuzzy c means clustering for profiling the customers who perceive the retail store based on its image. Fuzzy clustering is also compared with hard clustering solutions. Then the authors predict the model using discriminate analysis. Further, the chapter concentrates on the answer tree model of segmentation to identify the best predictor. Main conclusions with implications for retailing management are shown.

REPERIO

In this chapter, the authors describe Reperio, a flexible and generic industrial recommender system able to deal with several kinds of data sources (content-based, collaborative, social network) in the same framework and to work on multi-platforms (Web service in a multi-user mode and mobile device in a mono-user mode). The item-item matrix is the keystone of the architecture for its efficiency and flexibility properties. In the first part, the authors present core functionalities and requirements of recommendation in an industrial context. In the second part, they present the architecture of the system and the main issues involved in its development. In the last part, the authors report experimental results obtained using Reperio on benchmarks extracted from the Netflix Prize with different filtering strategies. To illustrate the interest and flexibility of the architecture, they also explain how it is possible to take into account, for recommendations, external sources of information. In particular, the authors show how to exploit user generated contents posted on social networks to fill the item-item matrix. The process proposed includes a step of opinion classification.