辅助古籍汉字分析的文献推荐方法分析word格式论文.docxVIP

摘要摘要古籍汉字结构复杂，种类繁多，参考文献版本众多、文献质量良莠不齐，使得古籍 汉字研究难度较大。针对这种情况，本文在分析古籍汉字研究过程之后，借鉴个性化推 荐领域知识，设计了一种辅助古籍汉字研究的文献推荐方法。首先，确定推荐方式。提取古籍汉字研究过程中的数字化信息，分析古籍汉字、研 究者、文献之间的关系，找到古籍汉字与文献之间的联系；分析推荐系统的优劣，选择 协同过滤推荐算法来实现文献推荐。其次，评分矩阵降维。根据古籍汉字逐步演变的特点，分析古籍汉字属性特征，使 用模糊隶属度方法计算古籍汉字字形模糊隶属度值。根据用户所研究的汉字找相似的汉 字，从而降低矩阵的维度。最后，设计文献推荐方式。通过分析用户心理来分析用户行为，设定用户行为评分 标准，使用隐式反馈评分方式记录用户行为，使用协同过滤推荐方式向用户推荐相关文 献。通过实验证明，本文所设计方法的准确度较高，推荐性能较好，有助于提高研究效 率。关键词汉字研究文献模糊隶属度协同过滤用户行为AbstractAbstractTextual Research on ancient Chinese character is difficult for many reasons. One of which is that ancient Chinese characters with complex structures have a variety of versions and references on this research which are a little patchy in terms of quality vary from types. Another is described as the pinch of researchers which result in a poor academic environment. Because of these reasons above, it have a difficult in making textual research on ancient Chinese character. The solution for this situation is to reference the recommendation system knowledge and design a kind of textual research method to implement literature recommendation.First of all, determine the method to recommend. Through analyzing the relationship among ancient Chinese characters, researchers and literature, we found the connection between ancient Chinese characters and literature. After comparing the effectiveness of a variety of recommendation systems, the filtering recommendation algorithm is chosen as the main literature recommendation method in this paper.Secondly, dimension reduction of score matrix. Calculate the similarity between ancient Chinese characters. Depending on the ancient Chinese character evolutional characteristics, use fuzzy membership degree method to calculate the characters’ fuzzy number value. According to the Chinese characters the users study on, the system finds the similar characters to reduce the dimensions of a matrix.Finally, design literature recommendation method. The method analyzes users’ behavior accord to users’ psychology to