Trust evaluation in virtual software development teams using BERT-based language models

Nowadays, people from different geographical areas can be closely related thanks to advances in information and communication technologies. This has a greater impact software development organizations where their members form virtual work teams. In these new co-located scenarios, the construction of interpersonal trust is more complex its very relevant performance paper presents results performanceevaluation four pre-trained language models based on BERT applied analysis tasks. For this work, small dataset 1453 comments obtained projects stored Github was created. The evaluated achieved moderately good values, order 0.84 for F1-score metric, which augurs that with further research they could significantly improved.