Application of Pca for context selection for collaborative filtering

Abstract

With the unprecedented expansion of the Internet, data overloading has become an increasing problem for retrieving useful information from internet. Users searching for products or content have endless number of Web pages to navigate and require enormous efforts, requires judgmental aptitude and intuitiveness to extract meaningful information from the enormous data. Context aware Recommender systems are meant to be an important solution to the data overload and skewed information problem that persists today in World Wide Web. One of the major challenges in the context aware recommender system is the selection of relevant contexts. The selection of a few most relevant contexts are important for accuracy in the recommender output, as irrelevant contexts decreases the accuracy of recommender output and also increases the computational complexity. There are various methods used for relevant context selection. Principal Component Analysis (PCA) has the advantage of dimensionality reduction that makes data processing efficient. Although PCA cannot be directly used for context selection, PCA can be used to extract the principal components those are linear functions of original contexts. In this paper, we propose an approach that does weighing of original features (contexts) from the principal components to determine the most relevant contexts. This approach is advantageous in terms of computational complexity due to dimensional reduction. © IDOSI Publications, 2014.