In this course project, I would like to explore the public discussion about farmers on Tanzanian Twitter. Agriculture is still the biggest employment by sector in Tanzania although with reducing trends from 83% in 1998 to 65% in 2019. This sector become the main economic activity for many Tanzanians, particularly those who live in rural areas. But Ochieng (2022) found that the agricultural sector in rural Tanzania is practised more by elderly people than by the youth. This condition implies that the jobs as a farmer seem unattractive for the young generation of Tanzania’s workforce. So, it is interesting to find out the discursive representation of farmers among Tanzanian social media users, which are dominated by the youth. Twitter was chosen because it is the more open and easy-to-access for social media analysis and become the second biggest social media by users in Tanzania with 22% of the population owning at least one account (Globalstats, 2022). Social media representation of the Tanzanian farmer and the main actors in the public discussion about it have huge implications to shape the wide public perception of working as a farmer in this country.

In this study, geotagging or geolocation is used to filter that the Tweet will be scrapped is only from the Tanzanian user accounts. This analysis can be conducted by social media text mining through the latent Dirichlet allocation (LDA) Topic modelling method. For the process of text mining to collect large textual data from Twitter, this study developed an algorithmic code by following the hashtag, symbol #. A hashtag works as index keywords or topics on Twitter. This function was created on Twitter and allows people to easily follow topics they are interested in. In addition, People use the hashtag symbol (#) before a relevant keyword or phrase in their Tweet to categorize those Tweets and help them show more easily in Twitter searches. This study will follow the hashtag of #farmer and #mkulima #wakulima (farmer in Swahili). But this quantitative method cannot solve the meaning-making for the result from social media text mining, so discourse analysis is used to interpret the results from topic modelling. Discourse analysis is essentially concerned with studying communication and meaning-making in context.