The multidimensional semantic model of text objects(MSMTO): A framework for text data analysis

Abstract

The modeling process for text-data type used for analysis purposes is to give a special representation for this kind of unstructured data. The given representation offers a formal description for text data to enable an effective use of the information contained in the text. In this context, and in order to perform analysis on this unstructured data type, we propose the multidimensional semantic model (MSMTO). The proposed model is based on the object paradigm. The model integrates a new concept Semantic Content Object used to represent and organize the semantic of text data in a hierarchical format, to enable a semantic analysis at different levels of granularity. Our modeling approach considers the internal composition of text documents as a structural hierarchy, which allows the user to perform analysis on different hierarchical levels. Our model offers also flexibility, by considering the semantic content of text-data as a measure, a fact or even a dimension.