Personal Data Analysis: Individual Examines 20 Years of Digital Communications to Map Friendships and Life Patterns

An individual has conducted an extensive analysis of their 20-year digital communication history, extracting and structuring approximately 1.2 million messages from multiple platforms to better understand their social relationships and emotional patterns. The project was motivated by a desire to move beyond traditional life-tracking methods—such as Tim Urban's "Life in Weeks" grid—to capture the qualitative aspects of relationships and personal fulfillment rather than just major life events.

The analysis drew from five major data sources spanning different eras of online communication: ICQ, IRC, and DC++ from the 2000s; VK, Twitter, and Facebook from the 2010s; and Instagram and Telegram from the 2010s-2020s. Using GDPR data access requests and similar legal frameworks, the individual obtained complete archives including messages, reactions, and social graphs. The technical work involved parsing and normalizing data from multiple formats—JSON and HTML files—while managing platform-specific quirks such as Instagram's double-encoded Cyrillic characters, Telegram's variable message IDs across exports, and Facebook's E2E encryption creating duplicate entries across folders.

After consolidating the disparate data sources into a uniform format, the analysis revealed significant patterns in communication noise and substance. In the longest conversation thread—486,000+ messages exchanged over ten years with a single partner—only 58.7 percent constituted substantive text, while 41 percent consisted of fillers, emoji-only messages, links, and media. The project aimed to build a personal Customer Relationship Management (CRM) system grounded in actual communication records rather than memory, enabling identification of patterns in emotional bandwidth, friendship cycles, and relationship half-lives that would otherwise remain invisible to individual perception.