An Improved K-Means Algorithm And Its Application In Bidding Data Analysis

Most of the Internet data are displayed in the form of text,which has great value.The efficiency of manual analysis is low and the accuracy is not guaranteed.As the increasing application demand of text value mining in various fields,the artificial intelligence natural language processing technology develops rapidly.There are a lot of bidding data on the Internet,which hides a lot of valuable information.Therefore,it is possible for enterprises to use text processing technology to automatically cluster bidding data and summarize data information automatically,and use the analysis results guide the production and operation activities of enterprises,help enterprises grasp the development trend of the industry and obtain the development opportunitiesAt present,there is a huge amount of bidding data on the Internet with different formats.The intelligent management and value information mining of bidding data requires a lot of manpower and material resources.Therefore,in response to the needs of clustering and automatic information extraction of bidding data,this paper uses natural language processing text mining technology to mine and manage its value.The main research content are as follows:Firstly,the research is based on improved K-Means' bidding information clustering algorithm.Aiming at the application requirements and main tasks of enterprise bidding information automatic clustering,research the bidding data preprocessing method,data vector representation model and similarity calculation method,and realize data preprocessing and data representation.Then,for the shortcomings of the original K-Means clustering algorithm,studying the initial clustering center determined method and the K value automatically determined method to realize the improvement of the K-Means algorithm,so as to realize the automatic clustering of the bidding data.Finally,use the bidding data set crawled from the Internet and the Python programming language to complete the experimental verification.Secondly,the automatic summarization algorithm of bidding information is studied.According to the requirement of automatic summarization of enterprise bidding information,this paper studies the data category extraction algorithm based on word frequency statistics to realize the automatic extraction of data category labels;studies the extraction algorithm of main construction content of bidding project based on rules to realize the intelligent extraction of main construction content of bidding project;and then uses the main construction content of data category label and cluster bidding data,According to a certain template,the summary information of each type of data is generated for researchers' reference;finally,use relevant software and bidding data to achieve experimental verification,and use the evaluation formula to evaluate the effect of the algorithm.