Recommender systems represent a prominent class of personalized Web applications, which particularly focus on the user-dependent filtering and selection of relevant information. Recommender Systems have been a subject of extensive research in Artificial Intelligence over the last decade, but with today’s increasing number of e-commerce environments on the Web, the demand for new approaches to intelligent product recommendation is higher than ever. There are more online users, more online channels, more vendors, more products, and, most importantly, increasingly complex products and services. These recent developments in the area of recommender systems generated new demands, in particular with respect to interactivity, adaptivity, and user preference elicitation. These challenges, however, are also in the focus of general Web page recommendation research. The goal of this chapter is to develop robust techniques to model noisy data sets containing an unknown number of overlapping categories and apply them for Web personalization and mining. In this chapter, rough set-based clustering approaches are used to discover Web user access patterns, and these techniques compute a number of clusters automatically from the Web log data using statistical techniques. The suitability of rough clustering approaches for Web page recommendation are measured using predictive accuracy metrics.
TopIntroduction
The page recommendation process based on Web usage mining consists of three phases: data preparation and transformation, pattern discovery, and recommendation. Of these, only the latter phase is performed in real-time. A variety of data mining techniques can be applied to this data in the pattern discovery phase, such as clustering, association rule mining, sequential pattern discovery, and probabilistic modelling. The results of the mining phase are transformed into aggregate user models, suitable for use in the recommendation phase (Bamshad Mobasher, et al., 2007). The recommendation engine considers the active user’s profile in conjunction with the discovered patterns to provide personalized content. From an architectural and algorithmic point of view personalization systems fall into three basic categories: Rule-based systems, content-filtering systems, and collaborative filtering systems. Our primary focus in this chapter is on model-based approaches to collaborative filtering in which models are learned through a variety of data mining techniques.
Rough sets theory is applied to design smart Computational Web Intelligence (CWI) systems. Rough Web Intelligence (RWI) has two major techniques which are
- 1.
Rough sets, and
- 2.
Web Technology (WT).
The main goal is to design intelligent rough e-agents which can deal with roughness of data, information and knowledge for e-Business applications effectively.
Rough set theory introduced by Pawlak (1982) deals with uncertainty and vagueness. Rough set theory became popular among scientists around the world due to its fundamental importance in the field of artificial intelligence and cognitive sciences. Similar to fuzzy set theory it is not an alternative to classical set theory but it is embedded in it. Rough set theory can be viewed as a specific implementation of Frege’s idea of vagueness, i.e., imprecision in the data is expressed by a boundary region of a set, and not by a partial membership, like in fuzzy set. The goal of this chapter is to develop robust techniques to model noisy data sets containing an unknown number of overlapping categories, and apply them for Web personalization and mining.
This chapter is organized as follows: the next section describes the background, followed by an explanation of the motivation behind this approach. Later, a description of the rough clustering approaches for Web page recommendation is covered. After this, the experimental results are presented, and the last section concludes this chapter.
TopBackground
Web mining is the use of data mining techniques to automatically discover and extract information from World Wide Web documents and services. Web mining is a technique to discover and analyze the useful information from the Web data. Web data mining can be divided in three general categories: Web content mining, Web structure mining and finally Web usage mining (Zaiane, et al., 1999). Web mining can be defined roughly as data mining using data generated by the Web and includes the following sub areas: Web content mining, Web usage mining, and Web structure mining (Srivastava, et al., 2000). In Web Content Mining (WCM) useful information is extracted from the content of Web pages (Pal, et al., 2002) as e.g. free text inside a Web page, semi-structured data such as HTML code, pictures, and downloadable files. Web Structure Mining (WSM) aims at generating a structural summary about the Web site and Web pages. While Web content mining mainly focuses on the structure of inner document, Web structure mining tries to discover the link structure of the hyperlinks at the inter document level. Web Usage Mining (WUM) is applied to the data generated by visits to a Web site, especially those contained in Web log files. Other sources can be browser logs, user profiles, user sessions, bookmarks, folders, and scrolls (Pal, et al., 2002).