基于特征项权重评估模型的英语试题关键词抽取算法研究  

Research on Keyword Extraction Algorithm of English Test Questions Based on Feature Weight Evaluation Model

作　　者：啜春红 CHUAI Chun-hong(Nanjing Vocational Institute of Mechatronic Technology,Nanjing 211135 China)

机构地区：[1]南京机电职业技术学院,江苏南京211135

出　　处：《自动化技术与应用》2024年第2期13-16,共4页Techniques of Automation and Applications

摘　　要：为提高英语试题数字资源的检索效率,提出一项基于特征项权重评估模型的英语试题关键词抽取算法。该算法主要由词同现因子、词长因子、位置因子、词频因子和特征项权重调节系数所组成。为提高该算法的通用性水平,研究还通过遗传算法对该模型的权重调节系数进行训练,并将权值优化后的算法应用于全国大学生英语四级考试试题的关键词抽取实验中。实验研究发现,研究所提出的特征项权重评估模型算法在各类英语试题的关键词抽取中均取得了良好的应用效果,平均准确率、平均召回率和平均F-measure三个方面的指标均高于TF-IDF算法和KEA算法,具有一定的应用价值。In order to improve the retrieval efficiency of English test question digital resources,this study proposes an English test question keyword extraction algorithm based on feature item weight evaluation model.The algorithm is mainly composed of word co-occurrence factor,word length factor,position factor,word frequency factor and feature item weight adjustment coefficient.In order to improve the generality of the algorithm,the weight adjustment coefficient of the model is trained by genetic algorithm,and the optimized algorithm is applied to the keyword extraction experiment of CET4.Through experimental research,it is found that the feature item weight evaluation model algorithm proposed in this study achieves good application results in keyword extraction of various English test questions.The indexes of average accuracy,average recall and average F-measure are higher than TF-IDF algorithm and Kea algorithm,which has certain application value.

关 键 词：英语试题 关键词 特征项权重评估模型 抽取算法 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]