Cognitive data analysis

• 主要作者: Chew, Jonathan Wei Liang
• 其他作者: Tan Ah Hwee
• 格式: Final Year Project
• 語言: English
• 出版: Nanyang Technological University 2020
• 主題: Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence
• 在線閱讀: https://hdl.handle.net/10356/139144
• 機構: Nanyang Technological University

As theoretical research into various fields, such as in the computer science field continues to grow, there is a high chance that newer research done would be more interdisciplinary in nature or consist of multiple research fields within the same discipline. It therefore becomes necessary for a researcher to be cognizant of the latest research trends for their research to be relevant. This project on cognitive data analysis, aims to analyze relevant data on computer science research conducted by Singapore Institutions in order to ascertain if there exists a prominent trend or pattern in computer science research in Singapore. The Fusion ART algorithm is used to perform clustering on the obtained data. ART algorithms have been shown to be performant when only a small number of datasets are available for training and is the reason for its adoption within this project. Firstly, a suitable dataset is chosen from a variety of data sources. Next, the collected dataset is preprocessed, through cleaning and transforming it into a form that can be utilized by the Fusion ART. We also evaluate suitable keyword extraction algorithms which would be used to tag a research paper to an appropriate ACM Computer Classification System category. Lastly, we experiment with the Fusion ART by varying its input parameters and feed the preprocessed data to the Fusion ART and examine the clusters formed and its relative weights in order to observe any trends within the dataset. Future work could be done to identify a better keyword extraction algorithm for tagging a research paper, as well as look into global trends in computer science research to identify computer science research trends on a global scale.