A Patent Recommender System based on Text Mining
| Autor: | Ou-Yang , Hau-Ching, 歐陽皓勤 |
|---|---|
| Rok vydání: | 2018 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 106 Patent retrieval is an important issue in patent management. How to filter out patents and other related patents that users might be interested in through information systems will effectively help users know how to develop patent map and avoid patent straps. In this research, we applied text mining technologies, including Chinese word segmentation, keyword analysis, word2vec, doc2vec, to develop a patent recommender system. The users only need to click on their current preferred patent documents, the recommender system can analyze approximately 1.78 million patents retrieved from national Intellectual Property Office. In advance, the similarity is calculated, and the highest relative patents can be filtered out. In the experimental results, the recommender systems which is developed based on Doc2vec shows the best performance. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|