Improvement And Application Of Slope One Recommendation Algorithm

With the development of the Internet,more and more people are enjoying the convenience brought by the recommendation system.Users can more easily obtain the interesting things through the recommendation system.The recommendation algorithm occupies an important position in the recommendation system.Slope One algorithm is a collaborative filtering algorithm based on items and is widely used in recommendation systems.Tourism is booming and smart tourism has become a new hot spot.The recommendation algorithm can accelerate the development of smart tourism.In the application,the Slope One algorithm often faces the problem of single weight and sparseness affecting the recommendation quality.In response to the challenges encountered by the Slope One algorithm,this paper proposes an optimization strategy for the Slope One algorithm,the main work is as follows:First,aiming at the Slope One algorithm using the number of users who have scored on two items at the same time are selected as the single weight of the score.This paper proposes a Multi-Weight Slope One(MWSO)algorithm,which provides positive help for predicting item ratings.Secondly,aiming at the sparseness problem,this paper proposes a Multi-weight Slope One with Auxiliary Item(AMWSO)algorithm.The AMWSO algorithm selects the items with the highest similarity to the predicted items as auxiliary items among the ungraded items of the user,and combines the user's ratings and the other user's ratings of the auxiliary items to obtain the auxiliary item scores,and combines the MWSO algorithm to predict the target item score.Third,the recommendation algorithm needs to pay attention to the problem of recommendation efficiency in practical applications.This paper proposes an Improved Multi-Weight Slope One with Auxiliary Item(IAMWSO)algorithm based on the AMWSO algorithm.By narrowing the selection of auxiliary items and combining multi-threading and cache technology,it can jointly improve time efficiency.Fourth,the IAMWSO algorithm is applied to the recommendation of tourist attractions.The crawler is used to obtain the original data of the website.After the data is cleaned,the IAMWSO algorithm is used to obtain the recommendation result.