Design And Implementation Of Big Data-based Free Travel Recommendation System

With the rapid development of Internet technology,more and more people starts searching and storing information and data from the internet.This makes the amount of data in the network showed dramatically growth.The Big Data age is coming,which means that people can access the data much more easily.But at the same time,it is more difficult to truly meet the individual information demands in the Big Data.Users are often lost in massive amounts of data,making them difficult to find valuable information for themselves.Although the continuous improvement of search technology allows us to quickly find the information through the search engine,depending solely on the search engine is difficult to examine and analyses for the user's further demands.To help users better find their own needs and to let the information of interest stand out from mass data,the personalized recommendation system based on Big Data analysis is introduced to help users quickly find useful information and takes the initiative to recommend to customers the contents that they are interested in.To help users better find their own needs and to let the information of interest stand out from mass data,the personalized recommendation system based on Big Data analysis is introduced to help users quickly find useful information and take the initiative to recommend to customers the contents that they are interested in.The recommendation system uses the recommended algorithm to recommend personalized information to the user based on the user's behavioral characteristics and interest preferences.This paper mainly studies the personalized recommendation technology based on the free travel big data analysis,which includes the content-based filtering technology and the collaborative filtering technology as well as the hybrid service based on the location service technology.The free travel recommendation system is designed for off-line,near-line,online three-part modules.The off-line part is designed for running time-consuming and large data calculation.The similarity modelling is used to solve the cold start problem of the recommended system.The system makes personalized recommendations for users based on their models through the online real-time processing.The system uses a hybrid recommendation algorithm to improve the recommended accuracy.The experiment results show that the recommended accuracy is effectively improved compared with a single recommendation algorithm.Free travel recommendation system uses a variety of big data processing platformtechnology.The system introduces the Hadoop platform MapReduce to process the data and classifies the data in parallel.Making use of the characteristics of Hermes platform ondemand data,system then loads the required data into memory for traversal search to improve the recommended system's ability to find and index large data.For different business scenarios,using one or more of the recommended algorithms for hybrid computing will improve the quality of personalized recommendation results.At last,the free travel recommended system test consists of the off-line experiments and online experiments.The MAE(Mean Absolute Error)and the comprehensive evaluation index are used in the off-line experiment.The training set and the test set are used to evaluate the quality of the recommended algorithm.On-line experiment uses the ABTesting verification method,and use the conversion rate indicators to evaluate the advantages and disadvantages of the old and new algorithms.At the end of this paper,it puts forward the improvement plan of the system and puts forward the anther's own view on the research direction of the free travel process recommendation system.