基于智能文本挖掘的技术水平预测研究——以声呐技术为例

Research on Technology Level Prediction Based on Intelligent Text Mining:An Example of Sonar Technology

[目的/意义]从开源文献中挖掘关于未来技术水平预测的观点信息,为我国前瞻预判科技风险、吸收利用国外技术专家智慧提供支撑。[方法/过程]设计“未来时间戳”词典,结合该词典和技术领域关键词,制定包含未来技术水平预测信息的文献检索策略。利用搜索引擎和学术文献检索平台获取文献全文,或批量下载文献关键字段信息(标题、摘要、关键词等)。利用模式匹配方法和“技术性能三元组”智能文本挖掘软件,从文献中抽取“未来时间戳”+“技术性能三元组”,并对不同国家、机构的预测结果进行总结归纳。[结果/结论]经过开展声呐技术领域未来技术水平预测实证分析,认为基于文本智能挖掘方法开展技术水平预测是可行的,且相对于趋势外推法、专家咨询法等具有结论权威、可靠、能够利用竞争对手国家专家智慧等优势。[局限]受限于文献全文的可获得性等因素,人工判读工作仍必不可少;在此基础上还可以进一步开展中外未来技术水平对比、技术威胁识别等研究工作。

[Purpose/significance]To study the method of mining the opinion information about future technology level prediction from open source literature.[Method/process]Design a“future timestamp”dictionary,combining the dictionary with keywords of technology fields,and formulate a literature search strategy that contains information about future technology level prediction.Utilize search engines and academic literature retrieval platforms to obtain the full text of the literature,or batch download the key fields of the literature(title,abstract,keywords,etc.).Using pattern matching method and“technical performance triad”intelligent text mining software,extract“future timestamp”+“technical performance triad”from the literature,and summarize the prediction results of different countries and different organizations.[Result/conclusion]After the empirical analysis of future technology level prediction in the field of sonar technology,it is concluded that it is feasible to carry out technology level prediction based on the text intelligent mining method,and it has the advantages of authoritative and reliable conclusions and the ability to make use of the wisdom of experts of rival countries compared with the trend extrapolation method and expert consultation method.[Limitations]Due to the availability of the full text of the literature and other factors,manual interpretation is still indispensable;on this basis,further research can also be carried out to compare the level of future technology between different countries.