Incremental Construction of Topic Hierarchies using Hierarchical Term Clustering

Topic hierarchies are very useful for managing, searching and browsing large repositories of text documents. The hierarchical clustering methods are used to support the construction of topic hierarchies in a unsupervised way. However, the traditional methods are ineffective in scenarios with growing text collections. In this paper, an incremental method for the construction of topic hierarchies are presented, allowing the update of a topic hierarchy without repeating the clustering process. The experimental results on several benchmark text collections show that our method obtains topic hierarchies with quality similar to traditional non-incremental algorithms.