Agricultural expert system - a participatory assessment

Abstract

Cyber Extension includes effective use of Information and Communication Technology, national and international information networks, Internet, Expert Systems, Multimedia Learning Systems and Computer based training systems to improve information access to the farmers, extension personnel and scientists. The dissemination of the technologies could be enhanced by using expert systems and other artificial intelligence technologies (Hadi et al., 2006).

An expert system is a computer-based program that uses knowledge, facts and different reasoning techniques to solve problems that normally require the abilities of human experts. The expert systems are based on the concept of artificial intelligence in which the experience and knowledge of human experts are captured in the form of IF-THEN rules and facts, to solve the field problems (Rao, 2003).

‘Diagnos-4’, was a computer-assisted software developed by Kerala Agricultural University during 2004. This package would support the agricultural extension workers and literate farmers for decision-making and help them in suggesting suitable control measures of the major pests and diseases of important nine crops of Kerala (Ganesan, 2002). It will be modified and released shortly for the benefit of all the stakeholders involved in agricultural development. Before introducing the system among users, it is appropriate to explore the possibilities of functioning of AES under the existing extension system so that suitable modifications can be made to make it more user friendly.

Development of AES, ‘Diagnos-4’ was the pioneering and ambitious programme of Kerala Agricultural University. The personnel involved in technology dissemination and technology users need much information on plant protection measures. Hence ‘Diagnos-4’ was selected purposively.

The research was conducted among the prospective users in two phases viz; exploratory design among researchers who were in the research institutes engaged in AES development and in TOT, all over India and experimental design among extension personnel and farmers from Palakkad District of Kerala. Mean scores, percentage analysis, Kendall’s Coefficient of Concordance, t-test for two samples assuming equal variances and Binary Logistic Regression were the statistical tools used in this study.

Twenty AES were identified during this study, developed by various agricultural research institutions in India. Many of the systems were restricted only to limited groups of users and they were yet to be popularized among the ultimate users. It was found that extension personnel and farmers possessed low level of knowledge especially in the areas of plant protection aspects of crops and they were in need of information on the same. Hence there is a lot of scope for the application of AES among extension personnel and farmers on plant protection aspects of crops that help the users to clarify their doubts, confirm their knowledge and provide real time information to the technology users.

Prospective users in the transfer of technology stream were very much satisfied about the future prospects of AES based on its better performance, settings in the AES, mode of presentation, practicability and serviceability of the system. The areas that needed modifications were: retrievability of information, relevancy of information and information content. Release of Malayalam Version with more emphasis on easy retrievability of information, needs the immediate attention of the researchers. All the categories of respondents perceived that AES had got ‘more potential’ in the transfer of technology in terms of disseminating information to the users.

The combination of AES and human expertise showed better performance and higher Information Efficiency Index (IEI) among the extension personnel and farmers. Majority of the extension personnel rated AES with high IEI. Whereas majority of the farmers rated AES with low IEI. Extension personnel and farmers assessed that the overall percentage of solution offered by AES in the plant protection of rice, coconut and banana was almost on par with the solutions given by human experts and in combination, it served better. It is better to introduce the AES designed separately for extension personnel and farmers. It is also necessary to release the software among the prospective users after a comprehensive orientation in using the AES. Maximum potential of AES can be explored by making the users as the partners in the AES development process to ensure user friendliness of Agricultural Expert System.