Design And Implementation Of OTA User Portrait Based On Big Data Technology

In recent decades,with the rapid development of the national economy,people's living standards have improved significantly.More and more people regard tourism as their first choice for holiday entertainment.With the increasing number of tourists and the accumulation of tourism data,it is both an opportunity and a challenge for OTA companies.If OTA companies want to survive better,improve their competitiveness,and obtain more benefits,they need to use massive data reasonably,analyze and mine user data,master user characteristics and needs,and provide users with accurate and personalized recommendations and Marketing services.In response to the above problems,this paper proposes a TF-IDF-LD-based multiple Naive Bayes algorithm model.This model improves the TF-IDF algorithm by introducing feature word position influencing factors and decentralized word frequency factors.A number of Naive Bayesian algorithms that have undergone Laplace smoothing process classify the data to generate user portraits.Finally,using big data technology,the OTA user portrait system based on TF-IDF-LD multiple naive Bayes algorithm is designed and implemented.The main relevant work of the paper is as follows:(1)In the study of user portraits,the naive Bayes algorithm with fast and stable classification is usually used for text classification.However,in the process of feature calculation and classification,the algorithm will automatically assume that all features have the same weight.This premise will ignore the characteristics of each text feature and reduce the accuracy of classification to a certain extent.In order to solve this problem,the TF-IDF algorithm improved by the feature word position influence factor and the decentralized word frequency factor is introduced,combined with a number of Naive Bayesian algorithms after Laplace smoothing to improve the accuracy of text classification Rate,applying it to the user portrait system can improve the performance of the system data processing module,and then improve the performance of the user portrait system.(2)Use big data technology to design and implement OTA user portrait system,use Flume technology to collect data,HBase technology to store data,use Spark,Hive,Presto technology to process data,and use improved algorithms for data classification and user portrait generation.Demand analysis of user portrait system,introduced the overall design of user portrait system.Finally,the design and implementation process of each module are introduced in detail,and the performance and function of the system are tested to ensure that the system can run normally.