Multi-dimensional text classification

Abstract

This paper proposes a multi-dimensional framework for classifying text documents. In this framework, the concept of multidimensional category model is introduced for representing classes. In contrast with traditional flat and hierarchical category models, the multi-dimensional category model classifies each text document in a collection using multiple predefined sets of categories, where each set corresponds to a dimension. Since a multi-dimensional model can be converted to flat and hierarchical models, three classification strategies are possible, i.e., classifying directly based on the multi-dimensional model and classifying with the equivalent flat or hierarchical models. The efficiency of these three classifications is investigated on two data sets. Using k-NN, naïve Bayes and centroidbased classifiers, the experimental results show that the multi-dimensional-based and hierarchical-based classification performs better than the flat-based classifications.

Classifying web documents in a hierarchy of categories: a comprehensive study

“…In the former case, the single-category assignment can be kept if the hierarchy is restructured. In the latter case, it would be better to consider a multi-dimensional framework, as that investigated by Theeramunkong and Lertnattee (2002). In the future, we intend to extend this work by considering the integration of both the multidimensional and the hierarchical frameworks, in order to support WebClass users with OLAP-like roll-up, drill-down and pivoting operations in an information retrieval context.…”

Image indexing in article component databases

“…An in‐depth description of how category labels were created for subject, time, and region facets is in Gelernter (2008). We limited the number of categories for each facet to between 10 and 15 because it has been shown that too many categories per level make browsing by category complicated visually (Theeramunkong & Lertnattee, 2002, p. 1).…”

Information management with multiple-categorization hierarchical structure

“…Due to the multi-dimensional characteristics of the data cube, they can apply multiple category structures, which lean to different aspects, to the searching results. [40] proposed a multiple-dimensional method to classify the text instead of applying flat and hierarchical structure, and proved that it was likely to attain higher automatic classification accuracy. Dennis Quan et al also implemented a multiple-categorization interface embed ded in browser and they prove that multiple-categorization are more flexible and natural to organize information from the aspect of human behaviors [36].…”

“…[18] and [40] emphasize on the internal structure and ignore the importance of user be haviors and user interfaces. [36], on the other hands, puts more efforts on the aspect of human behaviors and releases interfaces, which encourage people to multiply classify in formation.…”