Patent information aggregation and analysis

• Main Author: Chen, YanHua
• Other Authors: Anwitaman Datta
• Format: Final Year Project
• Language: English
• Published: 2014
• Subjects: DRNTU::Engineering::Computer science and engineering
• Online Access: http://hdl.handle.net/10356/59174
• Institution: Nanyang Technological University

With the fast advancement of technology, innovation ideas and creations need to be protected as a form of encouragement for research and development. There have been a numerous collection of patents filed over the last few decades. Through these patents, the project aims to analyze the information to identify emergence of trends and technologies. One is able to gain information and knowledge by visualizing historical data. In this project, a globally recognized organization, WIPO is crawled to extract the patent information that is available online. Furthermore, a categorization method has been developed to classify these patents into a hierarchy in the field of technology. The classified patents are then analyzed and compared to find patterns and trends. By comparing the results generated with Weka, it is found that unsupervised learning does not provide accurate class labels to categorize patents. The proposed method of classification is preferred. In addition, with the analysis of the data collected, trends can be found. One of the significant findings is the development focus in areas in network especially in network security and application development. There are some limitations to the research conducted due to the crawling difficulty caused by the target website. However, technologies relating to Android were analyzed and analysis could be drawn from the results. These results are able to certify many reviews and claims from other sources. Predictions and conclusion can be made through the analysis results.