A novel context-aware multimodal framework for persian sentiment analysis

Most recent works on sentiment analysis have exploited the text modality. However, millions of hours video recordings posted social media platforms everyday hold vital unstructured information that can be to more effectively gauge public perception. Multimodal offers an innovative solution computationally understand and harvest sentiments from videos by contextually exploiting audio, visual textual cues. In this paper, we, firstly, present a first its kind Persian multimodal dataset comprising than 800 utterances, as benchmark resource for researchers evaluate approaches in language. Secondly, we novel context-aware framework, simultaneously exploits acoustic, cues accurately determine expressed sentiment. We employ both decision-level (late) feature-level (early) fusion methods integrate affective cross-modal information. Experimental results demonstrate contextual integration features such textual, acoustic deliver better performance (91.39%) compared unimodal (89.24%).