Can music reflect the changing nature of sexual identity? How does the change in the sexual identity of LGBTQ+ representation be performed through music? To address these ques- tions, this thesis explores the changing performance of sexual identity through music lyrics. An extensive examination of LGBTQ+ representation in the music industry was conducted by compiling data from Spotify playlists. The sexual identities and self-disclosing dates of 54 selected artists were identified through manual coding, based on cross-checked news sources. Subsequently, the lyrics of all tracks by these artists were scraped from Genius. The collected data underwent various experiments including sentiment analysis, topic modeling, cluster- ing, and text classification techniques to observe changes in sentiment and topics expressed in the lyrics following the artists’ public disclosure of their sexual identities. Results show that: (1) the sentiment polarity could change in different directions depending on the specific characteristics of the artists; (2) artists express more confidence, more positive attitude, and stronger love after publicly disclosing their sexual identities. These findings provide empiri- cal evidence for the relationship between music and sexual identity and the impact of sexual identity on music expression, which addresses a gap in the existing literature by emphasiz- ing the connection between music and sexual identity. Furthermore, this study proves that music could not only reflect sexual identity but also reflect the dynamic change of sexual identity, which provides guidance for future studies in music and sexual identity. The use of computational methods could also provide some insights for future semantic analysis on a large corpus of lyrics.