This study develops and implements methods of identifying similar terms using collaboratively constructed folksonomies. In this study, two folksonomy-based methods are proposed with an aim of demonstrating the usefulness of folksonomy as a source for the discovery of similar terms, especially for 'non-in-the-dictionary' terms: co-occurrence-based and correlation-based methods. The experimental results show that the co-occurrence-based method performs comparatively better and that the folksonomies have a potential as a source for the discovery of similar or near-similar terms. The result implies that as the web2.0 service for the folksonomies evolves, the potential of folksonomy for the task will be increased.
展开▼
机译:本研究开发和实现使用协同构建的贩毒剂识别类似术语的方法。在这项研究中,提出了两种基于愚蠢的方法,目的是展示人物作为发现类似术语的源的源头的有用性,特别是对于“大字典”术语:基于共同发生和相关性基于基础的方法。实验结果表明,基于共生的方法比较较好,而且对人文的潜在作为发现类似或近似术语的源。结果表明,随着Web2.0为of of the the the toomies的服务发展而来,对于任务的潜力将增加。
展开▼