An Optimal Scaling Approach to Collaborative Filtering Using Categorical Principal Component Analysis and Neighborhood Formation

Abstract

Abstract. Collaborative Filtering (CF) is a popular technique employedby Recommender Systems, a term used to describe intelligent methods that generate personalized recommendations. The most common and accurate approaches to CF are based on latent factor models. Latent factor models can tackle two fundamental problems of CF, data sparsity and scalability and have received considerable attention in recent literature. In this work, we present an optimal scaling approach to address both of these problems using Categorical Principal Component Analysis for the lowrank approximation of the user-item ratings matrix, followed by a neighborhood formation step. The optimal scaling approach has the advantage that it can be easily extended to the case when there are missing data and restrictions for ordinal and numerical variables can be easily imposed. We considered different measurement levels for the user ratings on items, starting with a multiple nominal and consecutively applying nominal, ordinal and numeric levels. Experiments were executed on the MovieLens dataset, aiming to evaluate the aforementioned options in terms of accuracy. Results indicated that a combined approach (multiple nominal measurement level, "passive" missing data strategy) clearly outperformed the other tested options.

Visualizing clinical predictors of febrile neutropenia in Asian cancer patients receiving myelosuppressive chemotherapy

“…Moreover, the optimal scaling approach enables nominal and ordinal variables to be optimally transformed to variables with numeric properties, thereby overcoming the linear assumption that many data reduction techniques assume. 31 These features provide a flexible framework for us to work with for our analyses.…”

Determination of Recent In-Migrants in Yogyakarta using Bayesian Regression Logistics

“…Mixed scale indicators (metric and non-metric) used the non-linear principal component analysis to transform data. The non-linear transformation refers to the principal component analysis with optimal scaling from qualitative scale to quantitative value [13,14].…”

Movie Recommender Systems: Concepts, Methods, Challenges, and Future Directions

“…The K-means algorithm calculates the closeness of a feature to the centroid using the distance from the mean point. However, the principal component analysis creates a covariance matrix to calculate the eigenvectors and eigenvalues [42]. Therefore, it widens the scalability to find better comparisons to make the movie suggestions [43].…”

“…Data formulation: This is the first step where data is formulated and structured into tuples of dimension m × n. The possible tuples may depend on the user characteristics, the movie feature characteristics, or the combination of the user and feature characteristics [42]. The structure of the tuples is as given below in Table 2.…”