This project argues that news consumption from social media (Instagram) is fragmented and distorts how we understand the news narratives that are being shared by various sources. As a case study, the news topic of the ongoing anti-feminicide movement in Mexico and its social media content is used to explore how a narrative provides context for its fragmented representation. This research in progress creates new evaluation procedures for the detection of Mexican anti-feminicide movement narratives online using mixed-methods. The network science procedures and NLP/computer vision semantic models of the collected fragmented content provide social, legal, and ethical implications of detecting and monitoring online narratives of the Mexican anti-feminicide movement. Repository for dissertation on Social movement image narratives: https://github.com/lwdozal/SM_Narrative .