Summary

In this chapter, we introduced and appraised different approaches to recommendation systems. We examined the data structure requirements for content-based and collaborative filtering recommendation systems, and we discussed their underlining assumptions. We then point out the strengths of DataRobot in extracting features from challenging data types (for instance, image data) that normally limit the use of content-based systems. We then illustrated the use of DataRobot in building and making predictions using a content-based recommender system based on a small dataset.

It is important to highlight that the dataset used for this project was made up of multiple data types. DataRobot is capable of extracting features and integrating different data types to create ML models. In the next chapter, we will explore how to use datasets with a combination of image, text, and location data when creating ML models.