{"type":"video","version":"1.0","html":"<iframe src=\"https://www.loom.com/embed/7c0d5c8427084d759d30f8dfd07710cf\" frameborder=\"0\" width=\"1280\" height=\"960\" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>","height":960,"width":1280,"provider_name":"Loom","provider_url":"https://www.loom.com","thumbnail_height":960,"thumbnail_width":1280,"thumbnail_url":"https://cdn.loom.com/sessions/thumbnails/7c0d5c8427084d759d30f8dfd07710cf-21d105470c5681fd.gif","duration":221.499,"title":"Rwanda Content Recommender Demo, TF IDF Success 🚀","description":"Good afternoon everyone, I am Sultano, Kvlya Lemu, and my challenge ID is S2T13, a Rwanda content recommender. I built a TF IDF and cosine similarity recommender with local boost from feedback, supporting English, French, and Kinyarwanda searches. It returns the top five products made in Rwanda, with NDCG 0.988, local presence rate 100%, query time under 10 milliseconds, and evaluation showing 0.983 overall accuracy, 120 total queries and 400 total products. The system can send door to local artisans when new products are rapidly requested. No specific action was requested from viewers."}