Context-aware recommender systems for improved SME productivity

Abstract

SMEs in the UK are suffering from a productivity gap compared to larger companies and must find ways to maximise productivity in order to survive. With the widespread availability of digital tools, there is much choice for SME employees to take advantage of these to improve productivity. Tools can be adopted to improve a range of tasks and activities, such as, digital marketing, accounting, communication, etc. As a result, companies can improve productivity by positively impacting the rate of work, employee mental wellbeing, customer relationships, operational costs, and more. However, with the rapid increase in the number of digital tools on the market today, it is crucial that users are educated adequately on which tools to implement and how to utilise them. Context aware recommender systems can effectively learn about a user’s context and recommend items that would be suited to their needs. However, the context gathering process is key in determining the output. With this in mind, the research contributes an ontology-based context model (SMECAOnto) which gathers user context from SME employees such as, performance, emotions, and demographics. The context model is then used by proposed SME-CARS to determine a digital tool training intervention for users based on their needs with the aim of increasing effective adoption, and consequently, SME productivity. SMECAOnto is tested against competency questions through querying to test its effectiveness. The evaluation is promising and contributes a practical solution to the relatively understudied field of CARS and SME productivity.