A Scheme for Assessing the Usefulness of Business Video Reviews Based on Sentiment Analysis

As network technology continues to progress, an escalating multitude of business videos have surfaced. Among the myriad types of business videos, profit-oriented educational videos stand out as a significant category. Due to the proliferation of educational videos on online platforms and their increasing popularity and recognition, watching such videos has become a crucial means of learning for student users. However, discrepancies exist between certain courses, the pace of teachers’ lectures, and users’ requirements have contributed to a decline in course enrollment rates. Therefore, this paper takes the reviews under educational videos related to the sold courses as the object of study. Firstly, an LSTM model is used to analyze the sentiment of these reviews. Secondly, two factors related to the number of likes and follow-up reviews are quantified. Finally, by integrating sentiment tendencies with the number of likes and follow-up reviews, this paper establishes a usefulness evaluation scheme based on the LSTM model. The experiments demonstrate that this approach achieves high accuracy in the classification of review sentiment analysis and ensures the correctness of review usefulness scores, ultimately aiming to boost course sales rates.