Designing Hybrid Similarity based Search Engine Using Artificial Intelligence

Abstract

In the 20^th century it is been observed that there is drastic increase of information on web and the biggest reason for that is the availability of computation and content transferring on internet is increasing and it is not a less known fact that everybody is seeking for the most relevant information  , it is been already know that since the beginning of the era of internet the search is quite a challenging problem and also a necessary problem to solve  , however there are already plenty of solutions available for the search engines and they are serving us quite well but as the searching content and the users seeking for the content is increasing second by second it became necessary for us to move forward and experiment other techniques as well , best way to compare two sentences is to compare their similarities there are basically 2 types of similarities word based and semantic based similarity nowadays search engines are been built on either of similarity and in this paper we will see how can we implement a powerful search technique which leverages both the techniques , ideally semantic based similarity is a machine learning based technique which uses encoder model to generate semantic vector which are discussed in more detailed manner in the paper , another dimension that we shouldn’t ignore while solving this problem is time it is important to understand that user will spend more time .