In: Statistics and Probability
Personalization is a commonly used technique by search engines to return more customized search results based on individual users’ preferences. For example, if a user’s query is “apple” and we know that the user is a computer engineering, then search engines would return results relevant to the Apple product (such as Mac) rather than the pages relevant to the fruit. One possible way of capturing a search user’s preferences is to use his/previous queries and click-through information (i.e., which pages have been clicked for the previous queries). Please design a personalized search engine that can use the past queries and click-through information to personalize search results.
Discuss the pros and cons of your solution
Reference - Text Data Management and Analysis: A Practical Introduction to Information Retrieval and Text Mining, by ChengXiang Zhai and Sean Massung.