Finding Relevant Documents in a Search Engine Using N-Grams Model and Reinforcement Learning

The field of information retrieval (IR) is an important area in computer science, this domain helps us to find information that we are interested in from an important volume of information. A search engine is the best example of the application of information retrieval to get the most relevant results. In this paper, we propose a new recommendation approach for recommending relevant documents to a search engine’s users. In this work, we proposed a new approach for calculating the similarity between a user query and a list of documents in a search engine. The proposed method uses a new reinforcement learning algorithm based on n-grams model (i.e., a sub-sequence of n constructed elements from a given sequence) and a similarity measure. Results show that our method outperforms some methods from the literature with a high value of accuracy.