Content optimization is defined as the problem of selecting content items to present to a user who is intent on browsing for information. There are many variants of the problem, depending on the application and the different settings where the solution is used, such as articles published on portal websites, news personalization, recommendation of dynamically changing items (updates, tweets, etc.), computational advertising and many others. This work will address one variant that displays the best set of trending queries from a search engine in a module on the portal website. This application is different from the task of query suggestion in web search in the sense that it recommends popular queries to users from a certain pool of globally trending queries while query suggestion suggests queries relevant to what the user just submitted to a search engine. There are two major categories of approaches for content recommendation, content-based filtering, and collaborative filtering. The former one reflects the scenario where a recommender system monitors a document. A snapshot of Yahoo! front page. The page contains multiple recommendation modules such as Today module, Trending Now module, and News module. stream and pushes documents that match a user profile to the corresponding user. Then, the filtering system uses explicit relevance feedback from users to update the user’s profile using relevance feedback retrieval models or machine learning algorithms. Collaborative filtering goes beyond merely using document content to recommend items by taking advantage of information from other users with similar tastes and preferences. Previous studies have tried to combine both techniques for more effective content optimization.