The “follow” and “friend” mechanisms on social networking sites, combined with their widespread adoption, have led to the emergence of filter bubbles. Within these bubbles, individuals predominantly encounter opinions that reflect their own, thereby creating echo-chambers. In these chambers, ideas can become radicalized within isolated clusters, leading to societal divisions and potential confrontations. This study introduces SerendipitySeeker, a social networking service (SNS) viewer developed to address these issues. Some recommendation systems have attempted to include content outside typical user preferences to promote unexpected discoveries and mitigate personalization risks. However, according to the self-determination theory, individuals are more inclined to act based on intrinsic motivations than external directives. Considering this, we created interactive maps to position SNS messages on multiple axes, such as positive/negative and subjective/objective, using a language-model API. These maps are seamlessly integrated into an SNS viewer to broaden users’ informational horizons. Preliminary testing of SerendipitySeeker with 12 participants demonstrated its potential effectiveness. Users actively engaged with the map, shifting their content consumption from a linear pattern to a more comprehensive exploration.