Web环境下的科技新闻主题挖掘：以国际极地考察领域为例

Topic Mining of Science and Technology News in Web Environment: Taking the Field of International Polar Expedition as an Example

摘要:

科技新闻分析是科技情报服务的重要途径，将文本挖掘技术应用于科技新闻主题分析有助于快速高效地揭示科技发展动态。本研究提出一种基于LDA主题模型的科技新闻主题分析方法，从主题类型和主题强度角度进行主题演化分析。论文以2009—2018年中、澳、英、美4国极地科考新闻为数据来源进行案例分析。中文新闻中，极地测绘等主题的热度上升，极地冰川科考主题的热度下降；英文新闻中，热门主题为极地冰川科考与极地海洋科考；其余主题热度相对稳定。研究结果表明，该方法可以有效识别、抽取科技新闻主题并揭示其演化趋势，可以有效改善网络环境下科技情报分析的自动化程度。

Abstract:

Analysis of science and technology news is an important way of scientific and technological information service. The application of text mining technology to topic analysis of science and technology news is helpful to reveal the development trend of science and technology quickly and effectively. This study proposed a topic analysis method of science and technology news based on the LDA model and topic evolution analysis was conducted from the perspective of topic type and topic intensity. From 2009 to 2018, the polar expedition news of China, Australia, the UK and the USA were used as data sources for case analysis. In the Chinese news, the heat of topics like polar geomatics research rose but the heat of the polar glaciology research dropped. In the English news, the popular topics were polar glaciology research and polar oceanography research. The rest topics were relatively stable. The results showed that this method could effectively identify, extract and reveal the evolution trend of topics of science and technology news. It could effectively improve the automation degree of scientific and technological information analysis under the network environment.