Purpose: In this study, a content analysis of a platform that publishes content on the internet about blockchain technologies was made. The study aims to determine the factors (word and word string) affecting the reading rate of the digital content -on a titles basis- posted by the platform on Facebook. Method: 500 out of 2206 examples of content published between the specified dates were chosen randomly. The titles of the content were processed using standard text mining techniques and a new approach specific to the problem in this study on python programming language and then two different datasets were collected. The datasets were analysed using multiple linear regression. Findings: As a result of the analysis, it was discovered that some words and phrases used in the content titles affected the reading rate of the content. In addition, it has been determined that the new approach provides higher performance than standard text mining techniques. Implications: In this study, valuable information was obtained by processing raw data. As a result of the study, the theory was compared with the practice, and it was observed consistent results. It is determined that the new approach can be used effectively in similar text mining problems. Originality: The research relying on text mining was handled with a new approach on the basis of the content title. In this respect, the study has a unique quality.
Alan : Sosyal, Beşeri ve İdari Bilimler
Dergi Türü : Ulusal
Benzer Makaleler | Yazar | # |
---|
Makale | Yazar | # |
---|