[Purpose/Significance] Taking 14 provincial government open data platforms in China as the research object, this paper makes a comparative analysis of them from multiple dimensions, providing references and suggestions for the development of government open data platforms in China.[Method/Process] The crawler technology was used to acquire data, and the descriptive analysis of the data was carried out, and the Tf-idf model was used for text mining. Starting from the dimensions of data layer and platform layer, qualitative and quantitative analysis methods were used to compare fine granularity of data, domain distribution, timeliness, type of format, type of retrieval, access conversion rate and user feedback.[Result/Conclusion] At present, open data platforms in different provinces have different degrees of development, and there is certain room for improvement. For example, the release plan for data sets should take into account the province characteristic and the number of data sets, etc. In the process of construction, attention should be paid to the open platform data retrieval methods, training and user feedback.