基于模糊启发式的KNN算法在人才需求信息分类中的应用

摘要

随着当前计算机与移动互联网中数据的增长,在海量的网络招聘数据中如何有效挖掘可用信息成为当前教育和社会供需发展的技术瓶颈.为突破该技术瓶颈,提出了一种模糊启发式的KNN文本分类算法:基于ABC(artificial bee colony)的启发式搜索方法,以此来调整特征的权重,并利用模糊距离度量方法以测量测试观察和训练观察之间的相似性.先将招聘信息分词,利用TF-IDF(term frequency-inverse document frequency)算法与AP(affinity propagation)聚类算法进行特征选择和噪声数据剔除,最后采用结合启发式搜索和模糊距离度量的KNN 算法对文本信息分类.通过实验结果发现:该方法有效地解决了传统KNN 算法在人才需求文本分类方法中稳定性差和分类精度低的问题.%With the current massive growth of computer and mobile Internet data,in the mass of network recruitment data,how to effectively tap the available information has become the current education and social bottleneck in the development of supply and demand.In order to solve the technical bottleneck,a fuzzy heuristic KNN text classification algorithm is proposed:a heuristic search method based on ABC(Artificial Bee Colony)to adjust feature weights and a fuzzy distance measure method to measure similarity between test observation and training observation.Firstly,the recruitment information is segmented,and the feature selection and noise removal are performed by TF-IDF algorithm and AP(Affinity Propagation)clustering algorithm.Finally,KNN algorithm is combined with heuristic search and fuzzy distance metric to have Text information classification.The experimental results show that this method effectively solves the problem of poor stability and low classification accuracy of the traditional KNN algorithm in the text categorization of talent demand.