Abstract
The gig workforce, particularly platform workers, is experiencing a sharp increase worldwide, bringing attention to the systems that govern and evaluate them. This study examines how platform workers narrate their experiences with algorithmic rating systems using computational text analysis of publicly available YouTube interviews. Combining sentiment analysis, lexicon-based emotion detection, topic modelling, and keyword-in-context (KWIC) analysis, the research explores how workers describe earnings, work allocation, customer interactions, and performance evaluation. While overall sentiment across narratives appeared largely neutral, theme-level analysis revealed important variation. Earnings discussions were predominantly associated with anticipation and trust, reflecting future-oriented expectations of financial opportunity. In contrast, rating-related narratives showed higher levels of fear, sadness, and anger, indicating concerns around the opacity of evaluation and job insecurity. Customer-related narratives reflected emotional ambivalence, highlighting their dual role as income sources and evaluative agents. The findings demonstrate that earnings, ratings, and work allocation are experienced as interconnected elements of a unified performance system for platform workers. The study highlights that understanding how workers experience these systems is essential for advancing more transparent, fair, and sustainable models of digitally mediated employment.