IG entity tagging
Project: IG Entity Tagging (Outlier | Multimango) NER Annotation — Social Media / Visual-Linguistic AI Scope: The project involved annotating Instagram-sourced image-and-text content to support the training of multimodal AI models capable of understanding entities within informal, real-world social media contexts. The task required interpreting visual and contextual cues across approximately 40 annotated images. Data Labeling Tasks Performed: Identified and tagged named entities across five categories: people, locations, products, clothing, and informational entities within Instagram content For each tagged entity, located and matched an exact reference target image requiring visual search, comparison, and verification skills beyond standard text annotation Applied span-level highlighting to demarcate entity boundaries accurately within the content Quality Measures Adhered To: Worked within a structured review and rejection workflow, meaning submissions were evaluated against quality standards before acceptance Maintained consistency with annotation guidelines to minimize labeling errors and ensure inter-annotator reliability Exercised judgment in ambiguous cases (e.g. distinguishing between a tagged person vs. a brand, or a generic product vs. a specific item)