Loading...
Please wait, while we are loading the content...
Similar Documents
How You Can Identify Influencers in SAS® Social Media Analysis (And Why It Matters)
| Content Provider | Semantic Scholar |
|---|---|
| Author | Hatcher, Don Bawa, Gurpreet Singh Ville, Barry De |
| Copyright Year | 2011 |
| Abstract | There are many ways to calculate influence in social media. We use an approach that distinguishes between content generators and conversation generators. This allows us to distinguish between the generation of potentially passive content versus the generation of conversations that include content with “pass along” appeal. The “pass along” content is potentially more influential since it exposes the content to more actors and embodies a social “referral” component which has been demonstrated to affect the information and behavioral impact of messages. Content Generator Metric. In its simplest form, this is the number of communications over a given time period. A social media user who posts comments to a huge follower community generates a large content influence score. Unless the information is re-tweeted or mentioned in someone else's post, the “conversation” part of the influence score will be relatively low. Conversation Generator Metric. Social media users who drive conversations tend to generate a large number of file forwarding activities. They are mentioned a lot. They might also motivate a large number of replies and comments. Our original construction of metrics to operationalize these metrics used direct measures of tagged content in the social media source (in this case Twitter). Later, we experimented with the construction of metrics that used a standard network analytic framework. One nearly-universal framework uses 1 st and 2 nd degree inbound and outbound calculations to compute the metric. Our experimentation allows us to demonstrate that the original “custom” approach and the standard network-analytic approach are functionally equivalent. This confirms that our network-analytic approach to measuring influence taps into the notions of capturing both content and conversation. Further, that adoption of a standard approach enables us to extend our notions of measuring various aspects of influence – first developed in the Twitter Social Media data source – to other sources in Social Media. It also allows us to extend advances in social network analytics that are based on the inbound and outbound model to influence metrics that we build in our solution. In summary, our work strengthens our confidence and understanding of both custom and standard approaches, and provides additional insight to users of these metrics in our Social Media Analytics solution. |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | http://support.sas.com/resources/papers/proceedings11/319-2011.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
