Information diffusion has been studied through other life event models such as innovation spread, disease spread, technology adoptions etc.
The disease spreading model was been used to measure the rate and speed of the diffusion of the epidemic diseases. According to this model the propagation of the epidemic was based on the channel that exists in the network of hosts as susceptible, inflected, recovered, and last immune.[1]
The susceptible host is the first line of contact, while the immune is where the propagation ends. These are similar to social network nodes. The probability of infection decays away as the distance to the inflected increases.[2] Since these are nodes similar to the social network nodes, their role and impact in information diffusion can be measured likewise.
Generally, there are two models developed to model information diffusion.
· Horizon models-aim to capture the long-term changes (over the course of months, years, or even decades).
· Snapshot models- focus on short-term behaviour (weeks or months) while the background “chatter” topics are assumed to remain fixed.[3]
There are multiple factors that affect diffusion. The time any information spreads varies from system to system, from community to community and from a moment to other moments. For example “an S-shaped diffusion curve has been found for the majority of innovations adoption studied.”[4] This indicates that there are rises and falls in the innovation adoption processes. This diffusion model included five stages of the adoption process -awareness, interest, evaluation, trial and adoption.
When it comes to people there are five categories of adopters.
1. Innovators (first 2.5%),
2. Early adopters (next 13.5%),
3. Early majority (next 34%),
4. Late majority (next 34%) and
5. Late adopters or laggards (last 16%).[5]
The most important factor in this categorisation was the “invisible hand”- Social influence. This is where we can think of the importance of social influence in information diffusion. Two online connected nodes were found real neighbours in real social networks and their influence level at both setting were comparable. Social influence is developed out of different social and other statutes including culture relationships.[6]
The most important factor in this categorisation was the “invisible hand”- Social influence. This is where we can think of the importance of social influence in information diffusion. Two online connected nodes were found real neighbours in real social networks and their influence level at both setting were comparable. Social influence is developed out of different social and other statutes including culture relationships.[6]
Studies indicated that the sources of information also affects information diffusion and adoption. This supports the idea of social influence and the credibility or acceptance in a node or source.
It is believed that interpersonal influence is more efficacious than mass communication in bringing about social change. We can deduct from this that influencers in social media networks are important factors in communication.
Studies also indicated that information diffusion can be affected not only by the channel of communication and the context in which it propagates, but also by the type and content of the information itself. The following were found to be the typical contents that run through network channels.
· Spikes- sharp rising and falling topics in a network. This are for a short period of time but they get momentum very fast before sharp decline.
· Spiky Chatter- are the kind of topic many people chat about for more times. They show rise and fall (peaks and valleys) while they go through networks.
· Mostly chatter- these continuously move across networks for more times, slowly but steadily. [7]Spiky Chatter topics typically have a fairly high level of chatter, for example when communities are responding to external world events. Their persistent existence is what differentiates Spiky Chatter from Spikes. These topics are related to the two models we discussed above, though they do not occur separately. They can occur while people discuss the same topic. Thus, they can show the shift of people interest away from the discussion at hand, and thus can indicate sentiments, interests, etc.
The importance of measuring information diffusion
· Measuring content/ topic efficiency and effectiveness as related to time, goal and impact
· Deciding where and when to manipulate a topic or campaign to directly match to the users
· Measuring the uptake and adoption of newly introduced innovation
· Deciding and choosing the types of topics to post, high reaction topics such as the high resonance spiky chatter or low but steadily flowing topics[9]. To summarise, information diffusion is needed to study and document important topical issues such as abuse prevision, sentiment analysis and document matching. The knowledge can be applied to measure the overall effectiveness of communication systems, to make a real time data series analysis and information flow in social network.
