Besides sharing text messages, the direct and rapid dissemination of images and video has become increasingly common in social media. However, it has become less common for people to critically evaluate and confirm the authenticity of a message before receiving it. Previous research focused on text recognition or account authenticity but less focused on integrated multimedia promotional materials. In this article, we proposed a framework to recognize face data embedded with text in a message and deal with multimodal information. We combined Tesseract Optical Character Recognition (OCR) with Local Binary Pattern Histogram (LBPH) face training to recognize texts and faces. By extracting and analyzing pictures or posters circulated on social media, we identified and confirmed the information of a specific individual, as well as the text content related to it. We developed a system in which Tesseract OCR and LBPH face recognition showed an accuracy of 95.5 and 92.7%.