Text mining as an approach, referred to a text data mining, is the text analytics process of normalizing, integrating and grouping big data information including free (unstructured) texts and speeches unto the well-organized schematic models according to the research made by Cutting, 1992. Thus it is developed more in exploring text analysis, armed with Artificial intelligence, Natural learning process and other computational statistics applied in natural language as the mathematical approaches, Hearst (1999). Currently TM is used as the statistical tool approaches based on the key words which make mental model of the texts in discourse. One example is IBM’s Watson (Upbin, 2013). Further, with the support of TM algorithm, it is highly advantage of deriving important databases by its value from both texts and speeches with the methods such as sentiment analysis, clustering, figuring texts into images, macro structure and mental modelling in discourse of linguistic. Therefore, text mining is interdisciplinary manner approaches for data analysis, information processing, categorization in combination of mathematical statistics and linguistic analysis. To be collecting statistic databases, analyzing topic models, creating language model under the text mining approach, it brings us the specific characteristics of linguistics as text mining is to allow tounderstand the texts in different levels such as sentence, proposition, paragraph, context, idea of the whole discourse and to pre-calculate relations in between ideas and key and chosen words of the speakers. Text mining is also used for allocating, sub-parting and grouping the data into structured models of key concepts. Specifically, in linguistics, the text mining is importantly advised to release the same thought data the same as human language understands and processes. Another significance of the text mining, it works very well when variety different types of documents and data in various types including verbal and nonverbal through the language use are observed and created in semantic network which is fundamental upon to the topic concepts. The semantic network and nodes in key words is worthwhile to calculate correlations in between words’ meanings and values of the key expressions in the context of the discourse. With this advantage, this paper is explored the opportunities on how to parallel research on the political discourse and its global concepts and macro structures in the way the speakers choose the appropriate language and how they express their ideas, ideologies, political positions, implications, speech trends, and emotions in words-one hand, it is analyzed by the researcher’s view and other hand, NLP, ML, IA algorithms and mathematical statistics are used for collecting research data and mapping the superstructure and imaging global concepts and discourses-are in need of macro level analysis of discourse. And all the results from comprehensive method and TM approaches are re-proved according to the reliability of the analysis through the quantitative and qualitative analysis.
Read full abstract