Patent is an important carrier for recording the development of science and technology. It reveals the evolution trend of technology and research achievement of a country, and also provides valuable information that may notavailable in other scientific documents. The number of patent applications worldwide is rapidly increasing. To develop effective patent search systems is indispensable for inventors and professional patent searchers. When conductingpatent searches, the keywords provided by users are highly correlated with the accuracy of their search results. In addition, patent documents often contain a large amount of domain-specific knowledge and a wide range of technical terms. The content of patent documents varies in length and may be written in different languages, which increase the difficulty of the work of cross-lingual patent search and retrieval. As the number of patents filed in different countries continues to increase, the demand for searchers to use familiar languages to perform search tasks has also gradually increased, prompting scholars to begin to attach importance to cross-lingual patent search research. We plans to develop a crosslingual patent retrieval system in two-year project. In the first year, we aim to develop a patent retrieval system in a single language (i.e., English) based on the International Patent Classification (IPC) system and examine the effectiveness of the developed system. In the second year, a cross-lingual patent search system based on the Cooperative Patent Classification (CPC) will be developed. We plan to collect English patents from the United States Patent and Trademark Office (USPTO) and German patents from the German Patent and Trademark Office (Deutsches Patent-und Markenamt, DPMA). After the establishment of the patent retrieval system, we will invite the domain experts of the patent companies to conduct system validation, and use multiple evaluation indicators to make judgments to test the effectiveness of the proposed method.
|Effective start/end date||1/08/20 → 31/07/22|
- patent retrieval
- text mining
- patent mining
- cross-lingual retrieval
Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.