A scalable implementation model of mobile andragogy through data integration

Stephen Omondi Ogenga , Prof. Joseph Muliaro Wafula , Dr. Agnes Mindila

Data visualization, re-usable training objects, smart training

A scalable implementation model of mobile andragogy through data integration Stephen O. Ogenga*1 , Prof. Joseph Muliaro Wafula2 , Dr. Agnes Mindila3 *1. PhD Candidate, School of Computing and Information Technology - Jomo Kenyatta University of Agriculture and Technology (JKUAT), Nairobi, Kenya: email-omondiso@gmail.com/sogenga@nita.go.ke: 2. Lead Scientist- Africa Open Science Platform (AOSP) EA Node, Associate Professor, Department of Computing JKUAT Adjunct Professor, Africa Institute for Capacity Development (AICAD) Chair, CODATA Kenya, Chair CODATA-VizAfrica Series of conferences - Jomo Kenyatta University of Agriculture and Technology, Nairobi, Kenya 3. Director, School of Computing and Information Technology (SCIT) -Jomo Kenyatta University of Agriculture and Technology, Nairobi, Kenya Abstract Quality vocational education and training can be enhanced by technology to be more scalable, effective and efficient as indicated by (Latchem, 2017; Kirkwood et al, 2014). An observation by (Tradoc, 2017) indicates that progressive technological innovations allow the inclusion of a dynamic learning environment to challenge present and future citizens thereby maximizing their lifelong learning potentials. The use of appropriate technology allows facilitators to train individuals and teams at different levels of competence. Some technologies according to (Fominykh & Prasolova-Førland, 2012) allow for the use of a three dimensional (3D) collaborative virtual environments (CVEs). These CVEs enable advanced content manipulation, uploading, creating and sharing 3D objects and other media, such as text, graphics, sound and video. Content in this regard includes “objects, places, learning activities or other relevant experience”. Weak linkages between economic trends and occupational requirements have led to lack of timely information on the quality and quantity of skills that a developing economy needs. In Kenya, the informal sector holds over 80 percent of the country’s businesses and entrepreneurs characterized by ease of entry and exit, uncertified and low quality skill mix, reliance on indigenous resources, small scale operation, mobile financial transactions and Labour intensive technology. Enhancing productivity in the informal sector is an important variable since formal employment is not absorbing new entrants at the supply rate. As observed by (Weber, 2013; Hylen, 2015; Kamtsiou, 2016; Leibowitz, 2017), existing mobile learning platforms lack interconnection between the macro, meso and micro levels in the use of social media for vocational upskilling. This is buttressed by (Nomura, 2017) who observes that skills mismatches can occur at the macro-level, when the overall relevance of workforce skills does not trend with the demands of the Labour market. At the micro-level, it is noted that employers have to deploys expensive hiring processes to identify appropriately skilled candidates. When a country’s Labour market and education and training systems have marginal linkages then this is referred to as skills mismatch at macro-level. Labour market data can be of high volume, velocity and veracity hence a challenge to be displayed in real time. However, the massive data can be conditioned to provide Labour market intelligence according to (Laird et al, 2012; Maher et al, 2015). This can inform the supply and demand trends, emerging occupations and skill requirements in a Labour ecosystem. This chain can catalyze the profiling of job demands through an iterative pipeline process comprising four phases: characterization, extraction, comparison and updating. Properly visualized information can be interpreted by workers seeking opportunities and training providers interested in producing relevant and re-usable training content. As noted by (Wixom, 2016; Northern Research Consortium, 2012; SAS 2017), a society that is progressively digital, requires fast and real time delivery of visualizations to mobile devices while giving people the ability to easily explore data on their own. A smart training system is developed that enables the conversion of visualized real time labor market data into structured content for reusable training objects to enhance mobile skilling of informal sector workers. One of the key features of mobile cloud learning systems is their ability to deliver content and services on-the-go, enabling learners to engage with educational materials at their convenience (Leung & Chan, 2004). Moreover, these systems often incorporate contextual learning, where the learning experience is tailored to the learner's location, environment, and preferences. (Ruan et al., 2009) Additionally, mobile cloud learning frameworks have the potential to facilitate collaborative learning, allowing learners to interact and share knowledge with their peers, regardless of their physical location. (Wang et al., 2014). The proposed smart training system comprises three major parts namely; a real time labor market information processor (RTLMI) where data from labor market is converted to real time visuals to feed into a re-usable training objects module (RTOM). In this module, content developers create open and re-usable training objects anchored on Visual-Auditory-Kinesthetic (VAK) learning styles. The learning styles take cognizance of Multiple Intelligence (MI) theory which indicates that there are at least seven learning styles including Interpersonal, intra-personal, body/kinesthetic, visual/spatial, logical/mathematical, verbal/linguistic, and musical/rhythmic. Part three is a Mobile Training Content Interface (MTCI) where beneficiary trainees access the open training content on flexible skilling levels required. Open source data visualization for processing real time labor market information Real time Labour market data visualization is a key module in this system to deliver three components namely; stacked bar graphs, pie chart and heat map. The application is developed using HTML5, CSS3, PHP language and MySQL database management system (DBMS). Pilot secondary data from Informal sector training in Nairobi County is analyzed in terms of trade demands, youth placement and Master craftsmen distribution per sub-county in Nairobi county. The Lavarel PHP framework for web artisans used is an open-source software licensed under the MIT license. The features that enable this PHP framework to deliver web tasks more easily include; powerful dependency injection container; multiple back-ends for session and cache storage; database agnostic schema migrations and real-time event broadcasting. The pilot data is uploaded in csv format and properly formatted for precise results. The designed order of the attributes are; beneficiary name, youth training by gender, trade area chosen by youth and location of master craftsman offering training in that exact order. The demand for Trade Area Per Gender, per trade and per sub county are visualized using the horizontally stacked bar graph implemented using the Javascript library High charts. A pie chart is implemented using Javascript library Highcharts. The chart shows the percentage of demand per trade area in Nairobi county. The percentage per trade area = (youths placed in the trade area / total youths placed in Nairobi County) * 100 Heat Map This is an interactive GIS heat map created using QGIS, Leaflet Js, Leaflet-Heat Map JS plugin, Mapbox Imagery, and Open street map. The heat map component is used to compare the distribution of youth placement in the various sub-counties in Nairobi county. The map is made up of three layers; The base layer (powered by open street map and imagery by Mapbox); the polygon layer (made using QGIS and plotted using Leaflet js) and the Heat Map Layer (plotted using Leaflet JS and Leaflet Heat map JS plugin). The base layer loads and renders tile layers on the map from tile servers. Mapbox and Open Streep Map tile servers used to plot the base layer in the map. The Vector/Polygon Layer loads the polygon of all the sub-counties in Nairobi county. It is created using QGIS and in GeoJSON format for it to be rendered in Leaflet JS on the fly without having to use a plugin. The Heat map layer renders heat maps indicators in the sub-counties. The heat map layer is loaded using a leaflet-heatmap.js plugin and heatmap.js JavaScript library. Pilot heat map of informal sector training demands in Nairobi County An interactive GIS heat map created using QGIS, Leaflet Js, Leaflet-Heat Map JS plugin, Mapbox Imagery, and Open Street map. The heat map component is used to compare the distribution of youth placement in the various sub-counties in Nairobi County. The map is made up of three layers; The base layer (powered by open street map and imagery by Mapbox); the polygon layer (made using QGIS and plotted using Leaflet js) and the Heat Map Layer (plotted using Leaflet JS and Leaflet Heat map JS plugin). The base layer loads and renders tile layers on the map from tile servers. Mapbox and Open Streep Map tile servers are used to plot the base layer in the map. Fig.1 Pilot dash board showing heat map depicting demand levels for trades in specific locations in Nairobi County -2018 The dashboard in Fig 1 provides a visual representation of the skill demand locations of informal service providers where the red spots show the highest demands and the green spots are the second-level demand locations within Nairobi sub-counties. Re-usable training objects module (RTOM) This module provides content developers with an environment to create open and re-usable training objects anchored on Visual-Auditory-Kinesthetic (VAK) learning styles. The expected indicator is a percentage increase in the development of open data driven re-usable training content from visualized labor market information. It is designed such that content developers have the liberty to pilot regulated training content but consumers of the information have financial responsibilities for full access. Table 1 provides a summary of the elements of the proposed re-usable training objects template. Table 1: Re-usable training content development process Mobile Training Content Interface (MTCI) This is a platform where beneficiary trainees access the open training content on flexible skilling levels required. ThinkBoard software is used to create training content in light videos. Content is delivered on both web and mobile application pending copyrighting and trademarked as “FundiiView”. The verifiable indicator in this interface should be a percentage increase in the number of

Volume 10 Issue 1, January-2025

Pages : a569-a576