The evolution, spread, management and other relevant properties of business critical information in social media are vital for most businesses. Possible scenarios include containment: how one isolates harmful or not desired information spread, intervention: how one counters harmful information, transmission: how one introduces or infects the graph with new information so that its success can be estimated with reasonable accuracy, e.g. the case of competing memes.
The above scenarios are important for any business doing PR in social media. However, the methodology is not limited to social media only. One can use same methods in (m)any networked domain(s). For social media, our experience is that in order to get best possible results, one has to take a holistic approach: We have a set of subproblems that are not independent but there may be conditional independencies that can be exploited in the modeling process. The networked fashion of the problem helps in exploiting the conditional independencies. Overall this project is closely related to the work done in the project “Big Data Analysis in Social Media Applications”, and collaboration with the other team is actively supported. The biggest differences are issues related to multilingual data and meme evolution for specific business intelligence needs.