Research On Supply And Demand Matching Of Freight Platform Based On TF-IDF Algorithm

The current rise of freight matching platform allows more car and cargo resources to be stationed on the platform,the original intention of the car and cargo matching platform is to alleviate and eliminate the disadvantages from the asymmetry of information between the supply and demand sides,and to complete the resource integration plus transaction aggregation process.However,the current problems are:matching indicators need to be screened one by one,the satisfaction of both parties after completing orders is not high,the platform receives more complaints,the full load rate of vehicle owners is low,and the cost of completing transportation orders is high.In this paper,we study the development status of domestic mainstream freight platforms for vehicle and cargo supply and demand matching,describe in detail the problems related to cargo completion of vehicle and cargo matching,and elaborate the three modes and the whole process of vehicle and cargo matching.The model of LTL cargo matching is designed on the basis of comprehensive analysis of the indicators used by each vehiclecargo matching platform to measure the degree of cargo matching,and the collected vehicle source information and cargo source information are substituted into the model for instance verification to come up with a better matching scheme and test whether the efficiency of vehicle-cargo matching can be improved and the capacity resources can be used rationally.The main research work of this paper is as follows.(1)The development status of the vehicle-cargo matching platform is described,and three research directions related to the vehicle-cargo matching field are summarized.In order to filter to the appropriate matching index,the method of how to conduct index selection in related literature is compiled.In order to achieve a good supply and demand matching effect,the idea of building a supply and demand matching model is sorted out.The evaluation method of bilateral satisfaction is summarized,and the research content of TF-IDF algorithm is elaborated.(2)From the perspective of supply-demand matching,the core of freight transportation platform lies in developing a suitable matching model and building a complete matching process.By analyzing the current operation status of domestic mainstream freight platform,the complete operation process of three modes of bidding,order grabbing and order dispatching is summarized,and the complete supply-demand matching process of freight platform is sorted out by combining with the Yunman platform studied in this paper.From two perspectives of building matching model and setting intelligent solution algorithm,several common similarity measures are sorted out to prepare for the parameter solution of TF-IDF algorithm.(3)Quantify the satisfaction of both vehicle owners and cargo owners.The relevant data on satisfaction evaluation from the historical orders of the freight platform are inputted into the gray correlation model,and the three characteristics of relevance,similarity and similarity of the sequences are also considered,and the modeling solution is carried out sequentially to output the metric results of bilateral satisfaction,and the satisfaction is converted into satisfaction rate as the input data of the matching model.(4)Build the bilateral matching model for freight platform for supply and demand matching solution,and design the TF-IDF algorithm solution idea.Firstly,the respective supply and demand matching indicators of platform cargo owners and driver-vehicle owners are selected,and secondly,the indicator weights are determined by using the feedback competition method,and the cosine similarity of the vehicle source and cargo source indicator fields are calculated after processing the original data.(5)Some of the original vehicle source and cargo source data are collected from the Yunman platform,and the Py Charm software is used to encode the selected indicator fields and set the corresponding indicator weights respectively.In the first stage,the TFIDF algorithm is used to output the cosine similarity ranking of cargo sources,and in the second stage,a multi-objective optimization model is used to calculate in the dynamic matching process,and the vehicle and cargo supply and demand matching is carried out from four perspectives of comprehensive full load rate,satisfaction rate,yield rate and comprehensive matching rate,respectively,and the results are evaluated and compared,and the model and algorithm provide a diversified matching scheme selection for the owner users of the freight platform.There are 42 figures,25 tables,and 60 references.