There are two datasets related to our study on the relation between self-disclosure and online social well-being. Two separate files are provided; demographic variables and scale variables.

Netifier dataset is a collection of scraped posts on famous social media sites in Indonesia, such as Instagram, Twitter, and Kaskus aimed to do multi-label toxicity classification. The dataset consists of 7,773 texts. The author manually labelled ~7k samples into 4 categories: pornography, hate speech, racism, and radicalism.

A dataset containing customer commenters that obtain from user-generated content on Twitter and Instagram @byu_id. The dataset is used Bahasa, and it includes 32.684 raws. The following is an explanation of the variables in each column:

- comment: comments using Indonesian obtained from January 1 to June 30, 2021. This comment has been through a preprocessing process.

- sentiment: consists of neutral, positive, and negative sentiments of customers.

- dimension: there are six dimensions using customer experience proposed by Malviya and Varma (2012).

This study explores how male beauty influencers in Indonesia and Malaysia are redefining masculinity on Instagram, a platform that offers both opportunity and risk in culturally conservative societies. Through makeup, fashion, and aesthetic expression, these “celebgrams” challenge traditional gender norms while gaining visibility and building supportive online communities. Their presence is partly fueled by a strong female following and the “female gaze,” which helps reshape beauty ideals and opens up professional opportunities. While Instagram offers a relatively safe space for self-expression and empowerment, offline backlash reveals the tension between digital freedom and social conservatism. This research highlights how social media enables layered forms of identity negotiation and cultural change, offering insight into how gender expression is being reimagined in Southeast Asia.