Research on Automated Patent Map Construction

自动化专利地图构建研究

• 批准号: 17500063
• 负责人: YUKAWA Takashi
• 金额: $ 2.18万
• 依托单位: Nagaoka University of Technology.
• 依托单位国家: 日本
• 项目类别: Grant-in-Aid for Scientific Research (C)
• 财政年份: 2005
• 资助国家: 日本
• 起止时间: 2005 至 2006
• 项目状态: 已结题
• 来源: https://kaken.nii.ac.jp/en/grant/KAKENHI-PROJECT-17500063/
• 关键词: Patent classification; F-term classification; Patent map; Term weighting; Web application; Information retrieval; F-TERM; 概念ベース; ベクトル空間モデル

In this research, technologies for automated patent map construction are established.A term weighting classification method using the chi-square statistic is proposed and evaluated in the classification subtask at NTCIR-6 patent retrieval task. In this task, large numbers of patent applications are classified into F-term categories. Therefore, a patent classification system requires high classification speed, as well as high classification accuracy. The chi-square statistic can calculate the frequency of word appearance in the F-term and the frequency of word non-appearance in the F-term. The proposed method treats words as a scalar value and a ranking algorithm simply adds the word values of each word included in the test patent document in each F-term. Therefore, the proposed method provides classification that is significantly faster than other methods.The proposed method is evaluated in A-precision, R-precision, and F-measure. Although the proposed method did not obtain the best score, this method achieves a classification accuracy that is as high as those of other methods using machine learning or the vector classification method. In the NTCIR6 evaluation task, the processing speed is not evaluated. Therefore processing speed is evaluated on my own accord. The evaluation results show that the proposed method is much faster than that using the vector classification method.Evaluation results of classification accuracy and processing speed show that the proposed method is confirmed to be effective and to be practical.

本研究建立了专利地图的自动构建技术，提出了一种基于卡方统计量的术语权重分类方法，并在NTCIR-6专利检索任务的分类子任务中进行了评估。在这个任务中，将大量的专利申请归入F-Term类别。因此，专利分类系统既要求分类速度快，又要求分类精度高。卡方统计量可以计算F项中单词出现的频率和F项中单词未出现的频率。该方法将词视为标量值，排序算法简单地将测试专利文档中包含的每个词的词值相加到每个F项中。从A精度、R精度和F度量三个方面对该方法进行了评价。虽然本文提出的方法没有得到最好的结果，但该方法达到了与其他机器学习或向量分类方法一样高的分类精度。在NTCIR6评估任务中，不评估处理速度。因此，处理速度是我自己评估的。评价结果表明，该方法的分类速度明显快于矢量分类方法，分类精度和处理速度的评价结果表明，该方法是有效的，具有一定的实用性。

项目成果

Term Weighting Classification System Using the Chi-square Statistic for the Classification Subtask at NTCIR-6 Patent Retrieval Task

使用卡方统计进行 NTCIR-6 专利检索任务分类子任务的术语加权分类系统

• 发表时间: 2007
• 期刊: Proceedings of the 6th NTCIR Workshop Meeting
• 作者: Kotaro Hashimoto;Takashi Yukawa
• 通讯作者: Takashi Yukawa

A Synthesization of Multiple Answer Evaluation Measures using Machine Learning Technique for a QA System

使用机器学习技术对 QA 系统进行多项答案评估措施的综合

• 发表时间: 2005
• 期刊: Proceeding of NTCIR-5 Workshop Meeting
• 作者: Y.Matsuda;T.Yukawa
• 通讯作者: T.Yukawa