Car Recommendation System Version 1.0
Recommendation systems are taking more importance in online businesses, where the ability to propose a new item or product that a user will like can increase sales substantially. In this project, we propose to implement a web page where users can view certain types of items, for example cars, and give their feedback about them, either explicitly (thumbs up / down, likes, ratings...) or implicitly (clicking on item, spending time reading its description, sharing it..) Then, the system will run algorithms to come up with similar items to show to the user, and optionally collect feedback about the quality of the recommendation. Algorithms can range from simple similarity measures, to more complex machine learning models such as the SVM. Different topics of Data Science were applied, such as retrieving information, cleaning it for processing and storing it, evaluating algorithms for recommendation, and implementing big data processing pipelines. In other words, the system will study the user input throughout time, and recommend cars based on machine learning algorithms. This way the user can look into prospective cars that meet his or her expectations.