【摘要】 商品排名很大程度上促进商品的销量,如何提取优质的关键词是商品排名中的关键问题。传统的关键词提取算法基于词频或者词语间的相关性,忽略其他特征选项对关键词的影响。针对亚马逊平台商品信息特征,提出商品关键词提取方法 L-RAKE,结合RAKE算法及词语位置降低RAKE算法带来的误差,实验表明该方法具有更高的准确性。更多还原

【Abstract】 The ranking of products greatly promotes the sales of goods and how to extract high-quality keywords is a key issue in the ranking of products. The traditional keyword extraction algorithm is based on the word frequency or the correlation between words, ignoring the influence of other feature options on the keywords. Aiming at the characteristics of Amazon platform product text information, proposes a product keyword extraction method L-RAKE. The RAKE algorithm and the position of words are used to reduce the error caused by the RAKE algorithm. Experiments show that the method has higher accuracy.更多还原