基于概念格的Folksonomy知识组织研究——Tag Spam过滤指标权值配置

摘要

With the growing popularity and increasing reputation of folksonomy in practical applications, its problemsin the application is gradually revealed. The vicious tag and inefficient/ineffective tag spam to retrieve behavior of users seriously affect the results of practical application of folksonomy. Based on concept lattice theory and technology, this article first constructs the filter indicators system on tag spam according to the questionnaires and experts' intelligence, and constructs the concept lattice of tag spam filter indicators on the results of evaluation after the weight of tag spam filter indicators built in the traditional AHP method. In order to achieve the dynamically reasonable configuration on weight of tag spam filter indicators, the weight configuration of tag spam filter indicators is proposed validations with the association rules mining based on concept lattice, and the weights of tag spam filter indicators are dynamically revised according to result of validation. It attempts and explores to construct scientific, external project of weight configuration on tag spam filter indicators in folksonomy.%随着Folksonomy在实际应用中的日渐普及和声望的不断提高,其在应用中存在的问题也逐渐显露出来.恶意Tag和对用户检索行为无效或低效的Tag Spam严重地影响了Folksonomy的实际应用效果.以概念格的理论与技术为基础,首先根据问卷调查与专家智能构建了Tag Spam过滤指标体系,并在传统AHP法配置Tag Spam过滤指标权值基础上,针对评价结果构建了Tag Spam过滤指标概念格.通过基于概念格的关联规则的挖掘对Tag Spam过滤指标的权值配置进行验证,并根据验证结果动态调整相关指标的权值,从而实现Tag Spam过滤中指标权重的动态合理配置.对构建科学客观的Folksonomy中Tag Spam过滤指标权值配置方案进行了探索.