Text analytics on the case of Macedonian companies

Abstract

The subject of this paper is to present the content analysis with some of the latest text
analytics tools on the case of three Macedonian companies Alkaloid, NLB Banka and
Makedonski Telekom. The subject of text analytics is annual addressing of the CEO’s
(Chief Execute Officer) integral parts of the annual reports for 2016, 2017 and 2018.
Annual reports are used as a text input for the text analytics case. The text analytics
case is presented through bag of words and word cloud for determination of key terms
frequency. The frequency of the key word (tokens) selected in each category indicates
the focus of the company for the period addressed. At the same time this frequency
status suggests the focus of the company in the future involvements.
Sentiment analysis included in this text analytics case aims to determine the opinion
expressed by the CEO’s related to the reported and following business period. At the
same time feature of topic modeling is extracting topics from the CEO’s addressing for
each company for the observed period.
The goal of this paper is to present the analytical framework for business content
analysis and to test it on the text input in the example of the annual reports. In our case
results and findings from the analysis suggest that three companies through this
analytical framework present different management strategies with different focus on
operational and market segment. Alkaloid according to the result has orientation on
maximization of management performance and decision making. The focus on NLB
bank is on the continuous improvement on its competitive advantage against
competition. Although, the case of Makedonski Telekom with the presented results
suggested that the company focuses on the continuous improvement of services,
products and the network infrastructure supporting them.