基于实体词典与机器学习的基因命名实体识别

Gene Named Entity Recognition Based on Entity Dictionary and Machine Learning

将实体词典以特征的形式引入到机器学习模型中,提出一种基于实体词典与机器学习的基因命名实体识别方法,在GENIA 3.02语料上进行实验。测试结果表明引入实体词典特征后,在获得较高实体识别准确率的同时,优化CRFs识别模型的时间复杂度,提高系统识别效率。

By introducing the entity dictionary into the model of machine learning in the form of characteristics,this article proposes a method of gene- named entity recognition based on entity dictionary and machine learning and experiments on corpus GENIT 3.02.As indicated by the test results,after the characteristics of the entity dictionary are introduced,while a higher accuracy rate of entity recognition is obtained,the time complexity of CRFs recognition model is optimized and the systems recognition efficiency is enhanced.