Covid-19 pandemic is an ongoing global crisis which affects all the economic sectors including the most fundamental sector, agriculture. Bangladesh, a developing country and is home to 2.11% of the world's population, is facing a massive shock in the agricultural sector which is the backbone of the country's economy. In the era of technology and having lockdown in recent pandemic, every person willingly collects information on agriculture and rural economics from publicly available online news articles. To satisfy the readers' demand, not only the online blogs and newspapers, but also printed newspapers have taken a serious look at open access articles. This research analyzed entire online articles related to Covid-19 and Bangladesh agriculture news published during this country wide lockdown after the first case was addressed. Topic modeling algorithm has been used to determine the underlying topics over these textual contents for mitigating the gap between media coverage and actual scenario. This research considered the entire set of English online articles indexed by the most prominent Google search engine to point out diverse topics which are covered by various news sources. The dataset and experimental results were statistically analyzed to define the relationship and significance among different independent variables. Numerous interesting findings have been reported in this research including the fearsome number of female agricultural reporters, high frequency topic of agricultural economics and government spokesman, significance between underlying topics and article sources, etc. It concludes with recommendations to bridge the gap between government policy and actual scenario presented by news sources on Bangladeshi agricultural sector during Covid-19.
Keywords: Covid-19; agriculture news; content analysis; topic modeling; clustering algorithm
*Corresponding author: Email: moriom@sau.edu.bd
Tel: +8801520083662
