題目： Tech Mining Using Python
主講人：Scott W. Cunningham 教授
時間： 2016年5月11日9:00-11:00
地點：主樓418
主講人介紹：
Scott W. Cunningham，荷蘭代爾夫特理工大學政策研究及系統工程專業教授，技術管理領域著名期刊《Technological Forecasting and Social Change》雜志和《International Journal of Innovation and Technology Management》副主編，英國蘇塞克斯大學（University of Sussex）科技創新政策博士。曾在代爾夫特理工大學與哈爾濱工業大學的合作辦學項目任教，出版《Forecasting and Management of Technology》（第2版，Wiley出版社，2011），曾就職于美國電話電報公司及多家大型數據庫公司，從事以數據分析支撐決策指定的工作，并擔任電子產品制造業顧問，致力于通過決策方法研究、內容分析法、博弈論進行技術分析和戰略管理。此外，他還是Portland International Conference on the Management of Engineering & Technology (PICMET)和The International Conference on Innovative Methods for Innovation Management and Policy程序委員會委員，Technological Forecasting & Social Change, International Journal of Innovation and Technology Management, Scientometrics, PLoS ONE, Engineering and Technology Management等國際期刊的審稿人。
內容介紹：
In this lecture I provide a brief overview over tech mining, which is the process of measuring and instrumenting the innovation process. Tech mining is significant enterprise given the strong impacts of innovation on health, welfare, competitiveness and governance. Modern innovative processes involve understanding and anticipating spill-over effects, and the ability to anticipate and absorb the impacts of participating in an open, often global system of innovation. Open source movements – including open data, open innovation, and open source software – help to motivate a range of new approaches in tech mining. I provide a brief overview over a prominent model of innovation – the chain-linked process – and I describe some of the challenges of measuring innovative progress in the chain-linked model. Measurement involves using both input indicators as well as output indicators. Output indicators include intermediate measures of scientific or technological progress including publications, patents, and new product announcements. There are a variety of public and proprietary sources of information for use in tracking outputs. I describe some of the most prominent which I use in my own research. Tech mining is increasingly understood as a form of data science, with similar processes, techniques and methods. I provide a short overview over these processes and methods. In particular I describe the tools which are available to the tech miner. There are proprietary tools, including VantagePoint, as well as open source tools including the Python language. I discuss the strengths and weaknesses of both sets of tools. The lecture concludes with a demonstration of the range of data mining and transformation tasks available in Python. I discuss the resources available to help reduce the learning curve for Python data mining tasks.
（承辦：實驗室，科研與學術交流中心）
