Increasing public awareness of the environment has been the company’s concern for the last few years. According to Gray et al. (1987), companies growing public awareness in society can cause social problems related to pollution, resources, waste, product quality, product safety, rights, and labor status. The social control and critical action from the community encourage companies to communicate their environmental planning to all stakeholders, through sustainability reports (Modapothala and Issac, 2009; Nadhir, Wardhani, 2019).
Setiawan (2012) explains that sustainability reports are seen as a form of corporate social responsibility to stakeholders. Disclosure of sustainability reports can provide benefits for businesses both at national and global levels (Lawrence et al., 2013) because a company’s success is not only seen from its economic success (Shahin and Zairi, 2007), but companies must be skilled in balancing the interests of various stakeholders to gain a competitive advantage and have a credible reputation in the eyes of the public (Jamali et al., 2008).
Text Mining in a Sustainability Reporting
Data mining in the form of text where data sources are usually obtained from documents, and the goal is to find words that can represent the contents of the document so that the relationship between forms can be analyzed (Kumar and Bhatia, 2013). The text mining process includes information gathering, information retrieval, including link association and analysis, as well as predictive visualization and analysis. These characteristics are the same as the characteristics of a sustainability report, where the reporting style and format differ from one company to another. Modapothala et al. (2010) showed that based on the results of text mining, it was found that environmental variables were a factor that contributed more significantly to the explanation of sustainability reports. Goloshchapova et al. (2019) found that text mining results indicated that “employee safety”, “employee training support”, “carbon emissions”, “human rights”, “efficient power”, and “health drugs” were common topics reported by public companies in Europe and the UK. Freundlieb and Teuteberg (2013) also found that companies have broadened their understanding of CSR by referring to the term “sustainability” more often than “environment”.
Methods and Results
Research conducted by Iman Harymawan, Ardianto, Raden Roro Widya Ningtyas Soeprajitno, Melinda Cahyaning Ratri, and Yuanita Intan Paramitasari (2020) discuss Text Mining on Sustainability Reporting. The sample of this research is 152 sustainability reports of construction sector companies listed on the Indonesia Stock Exchange for the years 2010 to 2018. Reports are downloaded from GRI (Global Reporting Initiative) database. This study uses analysis techniques text mining using Python software. The study explains concern for sustainability issues as indicated by the tendency to express problems through negative words, while positive comments represent the company’s attention and efforts.
The results of text mining show that “sustainability”, “support”, “commitment”, “improvement”, “reward”, “integrated”, “fair”, “contribution”, “empowerment”, and “protection” are popular words related to company concerns and efforts to overcome economic, social and environmental problems reported in the sustainability report. In fact, “waste”, “complaint”, “disaster”, “corruption”, “dangerous”, “injury”, “conflict”, “violation,” noise “and” fraud “are popular words related to the problem. The main points disclosed in the sustainability report. Increased public awareness of the role of companies in society leads to increased social control and critical action in society. Therefore, companies must be able to give responsibility for the impact of their business activities on society through sustainability reporting.
Author: Iman Harymawan, Ph.D.
Harymawan, I .; Ardianto, A .; Soeprajitno, R .; Ratri, M .; Paramitasari, Y. 2020. Text mining on sustainability reporting: a case study, Journal of Security and Sustainability Issues 9 (M): 48-55. https://doi.org/10.9770/jssi.2020.9.M(4)