The traditional roles of a university are teaching and research with the aim of developing society and contributing positively to the national economic development by producing skilled and well-tutored graduates. Howe...The traditional roles of a university are teaching and research with the aim of developing society and contributing positively to the national economic development by producing skilled and well-tutored graduates. However, recruitments by these higher institutions are too reliant on the eligibility provided by Resumes of candidates, while neglecting their suitability drawn from their research activity and publications online. This study identifies insights in recruitment trends in higher institutions of learning and uses Artificial Intelligence to produce a more rounded and balanced decision-making process that caters for both eligibility and suitability. The methodology employs the machine learning process using the Multinomial Naïve Bayes for training the model as well as the Vader sentiment analyzer for accuracy and testing. The datasets used contained Resume instances as well as author publication information. The results show a score of 83.9% for the model as well as a sentiment analysis score of 1, indicating an overall positive score. The results show that sentiment analysis can help educational institutions in improving their recruitment models and attracting more suitable candidates for such roles.展开更多
文摘The traditional roles of a university are teaching and research with the aim of developing society and contributing positively to the national economic development by producing skilled and well-tutored graduates. However, recruitments by these higher institutions are too reliant on the eligibility provided by Resumes of candidates, while neglecting their suitability drawn from their research activity and publications online. This study identifies insights in recruitment trends in higher institutions of learning and uses Artificial Intelligence to produce a more rounded and balanced decision-making process that caters for both eligibility and suitability. The methodology employs the machine learning process using the Multinomial Naïve Bayes for training the model as well as the Vader sentiment analyzer for accuracy and testing. The datasets used contained Resume instances as well as author publication information. The results show a score of 83.9% for the model as well as a sentiment analysis score of 1, indicating an overall positive score. The results show that sentiment analysis can help educational institutions in improving their recruitment models and attracting more suitable candidates for such roles.