Organisations that invest in their people recognise the benefits of a skilled workforce. Even so, these investments are generally the first to be cut when there is a financial crisis. This is an important issue, especially as there is a strong call for life-long learning and the need to up skill with the rapid pace of technological advancements and societal changes. Consequently, more interest is being placed on evaluating these human capital investments to ascertain the return on investment. For example, the recent European Commission, Directorate General for Employment, Social Affairs and Inclusion study, titled “Skills pay dividends – accounting for human capital: how to increase employer training investment and make it more visible in company accounts”. However, these types of evaluations often prove to be challenging due to the intangible nature of the most immediate benefits: acquired skills and attitudes which must also translate into changed behaviours. This area is also challenging because of the lack of appropriate tools to assess whether the training is the reason for any change in behaviour.

This is a work-in-progress, and we are proposing a framework for leveraging the class of machine learning known as Natural Language Processing (NLP) to this end. It would not only lend scalability to the process, but – more importantly – accumulate indicators of these tacit changes during the natural course of work. Assuming essential compliance with ethical considerations, such as consent and data protection rights, managers could synthesize the corpus of online meeting transcripts, stored presentations and even email messages to appreciate changes more objectively in skills, attitudes, and behaviours before and after interventions. Aggregated results across teams could then be used when evaluating the impact of these interventions on the organisation’s KPIs. Such NLP analytics have already been applied to assess the effectiveness of cognitive behavioural therapy interventions in the domain of psychology. Our work should prove a novel adaptation of tools similar to IBM Watson and Linguistic Inquiry and in the human resource management domain.

Keywords: Human capital, impact of AI on education, natural language processing, learning analytics, training.