The invention allows organisations or companies to analyse the followers or fans on social media space, and identify a subset of high-valued audience who is highly possible to response and react on a specific topic of interested. A content-based approach is adopted to manage the dynamisms of the social space and not rely on any fixed ontology so that the latest and updated topics and entities can be extracted and used. This implementation requires minimal annotation efforts in identifying high-valued audience and it is applicable to any domain. In additional, the combination of recency and frequency analysis on top of the proposed content-based approach has shown promising result in identification of high value social audience.
This invention is able to dynamically generate a topic entity map based on various online resources. As this topic entity map developed is industry specific, it can also be used in other aspect such as emerging topic discovery. Clusters of social audience are determined based on contents published with consideration of temporal effect. The focus of the system is not on certain influencers but is capable of identifying audience who are highly likely interested in the topic or product. Value indexes are assigned to the social audience so that segmentation can be done and different incentive or engagement approaches can be applied for targeted marketing.
The system can be deployed as a service to companies or organisations who are interested to engage their fans/followers in social media space by providing them the insights to personalize their service or product to different clusters of fans/followers. Besides that, it can also be used in social media analytics and campaign provider for targeted marketing, to identify social audience who are more likely to response to a promotion or a product due to the specific topic.