Collaborative vs. Personalized Filtering

The web is becoming increasingly personalized in order to manage the vast amount of information one encounters on a daily basis. Websites have become attuned to how wary a consumer can become when browsing the web for information and have responded by engaging the consumer. A good example of the movement toward personalization is the travel industry online. Every travel website allows the customer to register with the website to facilitate future travel bookings. Once the travel company has the costumer’s information, they can send them emails concerning travel promotions that they would be interested in based on their home location or recent travel searches. This system establishes a personalized relationship between the consumer and the travel company, because the emails are like personal tips from a friend on how to save money. The relationship is equally advantageous for the travel companies as well because targeted marketing greatly increases their sales.
Collaborative filtering is another type of web personalization where the tastes and preferences of users are used to recommend items to other users with similar tastes, such as on the website Amazon.com. It is a system that assumes that if a certain customer chooses item A, then he would also like items B and C because other users who bought A also bought B and C. The flaw in this system is that it only increases sales for items that are already popular, leaving unselected items unseen by the customer. On the other hand, the system is effective in highlighting products that are of interest to the consumer. User ratings make the online shopping world a personal place by taking the existing system of consumption, which is normally driven by personal recommendations of products, and makes it an automated system on the web.