Recommendation Algorithm Based On Tag Time Weighting

There are two problems in Recommendation System: data sparseness and cold start. To deal with the problems of cold start and recommendation accuracy caused by the "information overload" of the computer network, the author came up with a recommendation algorithm which is based on tag time weighting The algorithm can effectively alleviate the cold start problem. First, the content and implementation process of the algorithm were introduced. Then, using the data sets, the proposed time-weighted label network structure algorithm, also called as TT-WN, was compared with the traditional collaborative filtering algorithm) and the classical network structure algorithm. Finally, the experimental results were compared and analyzed. The results showed that the proposed new time-weighted label network structure algorithm (TT-WN) did a good job in the experimental data sets, and the hit rate and sorting accuracy were much better than the classical network structure algorithm and the traditional collaborative filtering algorithm algorithms. Naturally, the conclusion that the TT-WN algorithm has many advantages was achieved.