Research On User Trip Behavior Based On Big Data Of Public Transportation

Against the background of rapid development of big data technology,governments at all levels have increased investment in urban transportation.For the construction of smart cities and the development of big data technology,in-depth analysis and mining of big data about intelligent public transportation are of great importance for governments' scientific management,commercial precision marketing and individuals' customized services.This paper starts from the research background and significance of the topic and the research status at home and abroad,and implements the public traffic big data analysis system based on big data technology and system design architecture.Based on the big data analysis system,it integrates multi-source data,studies user travel behavior,and builds user travel.Behavior model to mine users' potential travel behavior patterns.Mainly use the Hadoop big data distributed platform and relevant data stream processing engine tools to connect the data source to the non-destructive cleaning module,and then perform data preprocessing,including data quality assessment,data cleaning,and implementation of bus OD and subway OD algorithms for analysis.User complete travel chain.The Shenzhen area was divided into Shenzhen travel centers.Based on the Shenzhen POI data,the regional attriburte system was constructed.The regional attribute labels based on the TF-IDF weighted algorithm were extracted to analyze the POI attributes of the users.Combining the above work,this paper deeply analyzes the five travel dimensions of user travel time,user trip frequency,user trip site,user trip OD and user trip area,and implements a hierarchical user clustering algorithm based on the tripping site,according to the clustering result.All sites are divided into four functional business districts to study the potential preferences of users for business districts.At the same time,starting from the time dimension and space dimension,nearly 20 labels such as job-resident labels,fact labels,and predictive labels based on the K-means algorithm were extracted to construct a Shenzhen Tong users travel portrait system and tap potential consumers.Based on the user's commercial precision marketing and personalized customization services to provide a suitable business promotion program.