Mobile apps utilising AI for plant disease identification: A systematic review of user reviews

Plant pests and diseases are huge issues in agriculture which cause yield loss and threaten food security. There have been many efforts to educate, provide services and tools that help farmers deal with the issue of pests and diseases. Mobile apps that utilise Artificial Intelligence are one such tool that is being explored to have farmers identify diseases. There is evidence of mobile apps utilising Artificial Intelligence on mobile app stores; however, there is a dearth of studies showing how users perceive them. In this paper, we study the perception of users of these mobile apps by scraping 2438 user reviews from 9 Android mobile apps that employ AI. The average rating of the mobile apps is 3.48 and average the install is 1,000,000+. A thematic study on the user reviews is conducted and 19 concrete themes are discovered. The themes are divided across qualities that users liked, qualities they disliked and qualities that they requested. From the thematic analysis, the most liked quality was quality information on pests and diseases with 39 reviews. The second most liked quality being the correct identification of diseases with 29. The most disliked quality was the lack of plant coverage for disease and pests for identification with 135 reviews. This was followed by the issue of bugs with 66 reviews. Additional features were the most requested quality followed by more agricultural content.