| 英文摘要 |
The development of green energy technology is changing with each passing day. Although the traditional power industry still works well, various innovative power generating is emerging. In addition to the traditional power industry, various technology companies and new startups have also joined the green energy industry competition, and they also use patents to protect innovative technologies. The number of patent applications related to green energy has hit a record high. It has become an important issue that how to analyze the huge amount of data from these patents. This study uses artificial intelligence natural language processing technology to assist in analyzing the huge data of green energy patents. First, the word segmentation tool is used to help the computer understand the related vocabulary of fintech patents, each vocabulary is assembled into phrases, and then the semantic role played by each phrase is marked, and then machine learning is used to make the computer automatically process each sentence to generate appropriate structure tree, and the algorithm based on Bert model, allows the computer to read a large number of patent documents, convert vocabulary into ''word vectors'', and use a large number of patent document training to strengthen the computer's abstract thinking ability. When encountering related green power patent literature for new technologies or applications in the future, it will also be able to understand its content. In the current preliminary research results, artificial intelligence algorithms have been used to identify mobile platform-related technologies in green energy patents, and can target different applications, such as wind power, solar power, and geothermal energy preliminary classification and analysis of patents of different applications. This study finds that the use of artificial intelligence technology to analyze such huge amounts of data has great potential, and can provide appropriate guidance for green energy patent applicants in patent layout. |