An Extended Cognitive Situation Model for Capturing Subjective Dynamics of Events from Social Media

The event-indexing situation models are introduced as event models derived from language to facilitate comprehension and memory retrieval. These models explain how fragmental information about events are collected, integrated and updated into a coherent set of views of what the text is about. The models are adopted as the basis of this study as an attempt to capture the event with contextual, dynamic, and social features, as conveyed by the vast volumes of online textual resources. Information in social media is received through highly personalized channels and is reshaped and interpreted in a more individual, segmental and real-time manner. The reprocessed information is then spread at high speed to a wider range of receivers. Therefore, the interpretation of mainstream media content is influenced by ongoing and dynamic contribution of opinions by users empowered by social media. This new phenomenon has not been examined so far from the perspective of the impact on conventional situation models. This paper explores how collaborative and sharing aspects of social media emphasize subjectivity of interpretation of mainstream media and proposes an extended cognitive situation model which better represents event-centric knowledge. This study investigates the mechanisms for constructing and updating the situation models with continuous textual information streamed from heterogeneous forms of media. It also empirically demonstrates how the proposed model can enhance the understanding of subjective aspects of events with dynamic social opinions