摘要
随着法律文本的大量公开,在文本挖掘中发挥越来越重要的作用,同时随着机器学习与自然语言处理技术的发展,其与大数据的融合运用逐渐普及,将两者结合起来进行相似度判别分析,有利于对法律文本的充分挖掘利用,对于帮助用户了解案情,同时基于法律文本相似度分析可拓展更多应用,对于促进国家法制化建设具有重要意义。使用了一种基于机器学习的相似度算法,通过与不同词向量结合的方式,能够有效提升文本相似度对比的准确率。通过在真实民间借贷类法律文本比较案例中实验取得明显效果,准确率提升10%,具有较好的使用前景。
With the large number of legal texts being published,they have played an increasingly important role in text mining.At the same time,with the development of machine learning and natural language processing technologies,the integration and use of big data has gradually become popular.Similarity calculation analysis is conducive to the full mining and utilization of legal texts.It is helpful for users to understand the case.At the same time,it can expand more applications based on the similarity analysis of legal texts.This paper uses a similarity algorithm based on machine learning,which can effectively improve the accuracy of text similarity comparison by combining with word vectors.The experiment has achieved obvious results in a comparative case of real civil lending legal texts,and the accuracy rate is improved by 10%,which has a good prospect for use.
出处
《工业控制计算机》
2020年第6期3-5,共3页
Industrial Control Computer
关键词
民间借贷
法律文本
相似度计算
机器学习
自然语言处理
private loan
legal text
similarity calculation
machine learning
natural language processing
