Effective mobile query systems for rural farmers

Crynodeb

This thesis considers the appropriate arrangement and access of knowledge to help rural farmers in Nigeria improve the ways in which they produce crops and raise livestock. A review of the trends in mobile computing-based extension delivery in Nigeria was conducted. The review showed that there is a high penetration of mobile telephony in Nigeria, but agricultural support systems have not been adequately developed to take advantage of this mobile phone coverage. One of the major challenges identified was that although farmers have access to mobile phones and the Internet, they find it difficult to access information easily and quickly to address their local farming needs. This led to the exploration of effective ways of facilitating farmers’ access to locally relevant information. Requirements for effective mobile farm query systems were developed and established. They include (1) Minimal input by the user (2) Effective selection of information relevant to the user (3) Restricted access to the Internet (4) Wide range of help for farmers (5) Easy expansion of the system (6) Availability in minority languages and (7) Sustainability within developing country budgets. A mobile query technique that meets the identified requirements of providing access to information for rural farmers by ontology-based partitioning was developed. The technique requires the repeated partitioning of the domain so that the user can select the appropriate partition at any level, and gradually home in on the desired answer. This technique has been used to develop a farming advice system for rural farmers in Jos, Plateau State, Nigeria. The farming advice system partitions the farming advice domain into three major groups i.e. Grains/tubers, Vegetables and Animals. These groups are further partitioned into sub-groups based on how local farmers group farming concerns and with some inspiration from case based reasoning. Question and Answer (Q/A) pairs were collected from local small holder farmers with the active collaboration of local agricultural experts from the Federal College of Animal Health and Production Technology, (FCAH&PT), Vom, Plateau State, Nigeria. The local agricultural experts conducted focus group discussions (FGDs), one-to-one meetings and questionnaire surveys with the local farmers. This information was curated, and the mobile query technique was used to arrange the Q/A pairs in the farming advice system. Farming advice for fifteen species of animals and crops were captured in the mobile farm query system. It was observed that within an ontological partition, similar sub-partitioning occurs e.g. for crops - sowing, harvest, storing, fertilizer, general production, pest/disease and weeding. Animals - Breeding, Disease, Feed/water, general production and vaccination. As more crops and animals are added to the system, the set of concerns can be used as a guide in the collection of Q/A pairs from local farmers and extension workers. The application of the mobile query technique to other domains was explored by considering how it could be used for making FAQs about Aberystwyth University Information Services accessible. A third domain explored was the provision of personal health advice. The three domains used in implementing the mobile query technique are very different but the developed principles for breaking down a domain have proven effective in each. The mobile farm query system developed using the technique is being trialled by small holder farmers in Plateau State, Nigeria. The system provides small holder farmers with an easy and timely way of accessing farm query information. This is very important to farmers because of the critical nature of timely information. The system provides extension agents who may not have immediate answers to all farmers’ queries when on the field quick access to this information even in areas where there is no mobile signal or Internet access. It also provides a database of farmers’ problems and concerns. As farmers use the system, the frequency of their consultation of Q/A pairs, the season when these Q/A pairs are consulted and the location from where they are consulted can be used to build a big picture of farming problems and concerns within an area. This information can be used to properly channel much needed help to small holder farmers

Dyddiad Dyfarnu	2020
Iaith wreiddiol	Saesneg
Sefydliad Dyfarnu	Prifysgol Aberystwyth
Goruchwyliwr	Chris Price (Goruchwylydd)