KNOWLEDGE GRAPH BONDING WITH THEMES AND CHARACTERS: AN EDUCATIONAL DATA MINING STUDY

Abstract

A close reading of big data is a time-consuming activity, then finding the interrelationship of themes and characters is a more challenging task for social science learners. To address these challenges, Links tool draws knowledge graphs that show linked open data as a precise visual. These knowledge graphs look like multipronged links in the Educational Data Mining process, and they are interactive in nature to fulfil the needs of digital natives. Links tool constructs knowledge graphs without any programming skill; hence, this data visualization has become very beneficial for learners and teachers during the pedagogical phase and knowledge discovery. This study is an experiment of Links tool on five biographical essays. The current study employs mixed methods for data generation; Connectivism Theory and Hermeneutica Theory have been applied during data analysis. Furthermore, zooming and filtering techniques have also been applied to comprehend the true hermeneutic meanings of Knowledge Graphs. Major findings of this study reveal that knowledge patterns have been extracted in the form of a correct drawing of knowledge graphs, thus, they serve digital hermeneutic purposes. Moreover, bonding of characters and major events show 88.88% accuracy when compared with the source text; hence they reveal epistemological and ontological characteristics.