Next: Effect of Similarity Algorithms Up: Experimental Evaluation Previous: Experimental platform.
In this section we present our experimental results of applying item-based collaborative filtering techniques for generating predictions. Our results are mainly divided into two parts-quality results and performance results. In assessing the quality of recommendations, we first determined the sensitivity of some parameters before running the main experiment. These parameters include the neighborhood size, the value of the train/test ratio x, and effects of different similarity measures. For determining the sensitivity of various parameters, we focused only on the train data set and further divided it into a train and a test portion and used them to learn the parameters.
- Effect of Similarity Algorithms
- Sensitivity of Train/Test Ratio
- Experiments with neighborhood size
- Quality Experiments
- Performance Results
Badrul M. Sarwar