User Adoption Analysis of Social Media Applications Based on NLP and Technology Acceptance Model - A Case Study of Facebook

With the growth of social media apps and the regular release of new apps, it is crucial from a business analytics perspective to understand whether users accept newly released mobile applications and to analyze the factors that affect user acceptance. This study presents the first analytical model that integrates natural language processing based on user reviews with Technology Acceptance Model to address this issue. By processing user reviews in the app store with sentiment analysis, the sentiment score corresponding to each review is derived and considered as quantitative data for the structural model. Afterward, for each user review with sentiment score, a TAM keyword lexicon extended by WordNet is used to classify the user acceptance indicators and processed into vectors of the same length according to the time series. Finally, regression estimation of the structural equation model was performed using AMOS. The model results showed that out of all seven hypotheses, five were supported and two were rejected. The most significant factor affecting users' adoption is user attitude. And perceived ease of use and perceived usefulness have a positive impact on users' attitudes. In addition, satisfaction is also one of the factors influencing user adoption.