Search engines play a vital part in everyday life by assisting users in finding information relevant to a set of keywords. Since more information are released on a daily basis, the abundance of data has provided a significant challenge to current search engines in terms of locating the most relevant information to a given query. While  search engines frequently return a huge number of relevant articles, users must also extract meaningful information from the retrieved articles. Taking the relevant information from the articles and presenting it in a user-friendly fashion will optimize the search engine advantages. Search engines employ particular match strategies on Inverted Index to help obtain a small collection of relevant documents from billions of web pages in order to achieve excellent result quality and low query response time. In this report we hope to improve the search response using deep reinforcement learning by comparing Parameterized Action Soft Actor-Critic (PASAC) to various other deep reinforcement learning algorithms. These findings will then be compared to each other and Lexical Search to see if these algorithms yield relevant results.