Application Research of Naive Bayes Algorithm Based on DIKW in Weather Website

EasyChair Preprint no. 5685, version 4

Abstract

The generation and update of weather website tourism products provide tourists with a reference for the weather conditions of the destinations. However, due to various reasons, they cannot be updated on time. Manual monitoring is required to update. If there is no update, manual update is required, which undoubtedly increases the burden on business personnel. The intelligent age needs to find a time-saving and labor-saving solution to solve this problem. The DIKW (Data Information Knowledge Wisdom) model is the most basic model in the research of information management, knowledge management, etc. The main content of the weather website is the collection of historical data, the summary storage of information, the knowledge processed by machine learning, and the wisdom of decision-making applications. The Naive Bayes Classification Algorithm (NBC) is widely used because of its high classification accuracy and simple model. For this reason, the model is combined with machine learning algorithms and imported into actual meteorological application research, and the NBC algorithm is combined with the weather website. One-month historical update data mining calculates the prior probability, and then the classification result is calculated according to the Python program to capture the data of the day. By recording the future 16 sample data sets using this model for calculation and analysis, 15 pieces of data conform to the results of the model calculation and classification, and the accuracy rate reaches 93.7%. The results show that the higher accuracy of the algorithm classification forecast can promptly remind business personnel, and better guarantee the timely update of tourism products, thereby improving the work efficiency of business personnel, and providing practical application reference value for smart weather service business automation.

Keyphrases: classification prediction, Crawler, meteorological, Naive Bayes, weather website

@Booklet{EasyChair:5685,
  author = {Chaoning Li and Liang Chen and Shenghong Wu and Yunyin Mo and Liying Chen},
  title = {Application Research of Naive Bayes Algorithm Based on DIKW in Weather Website},
  howpublished = {EasyChair Preprint no. 5685},

  year = {EasyChair, 2021}}