Authoritative Prediction of Website Based on Deep Learning

Website authoritativeness is generally measured by external links, the more high-quality external links, the higher the authority of the site. Algorithms such as PageRank, etc. are com-monly used in the evaluation of website authoritative, However, this kind of algorithm is selective to evaluate the authoritativeness of websites, which makes this method have some deficiencies. This paper uses deep learning method to evaluate the authority of different websites under a certain search query by mapping the search query and the corresponding title of the website as a vector and calculating the similarity between two vectors. Websites of high results are referred to as authoritative sites under the search query, thus providing a new perspective to measure website authoritativeness. By comparing the three model experiments with Word2vec, CNN and LSTM, the experimental results on open datasets show that it is effective to use these three models, of which the LSTM model works best.