Database Reference
In-Depth Information
Marital status Whether the poster is married
Employment status
Whether the poster is employed
Workflow: Pulling It All Together
To successfully assess what people are saying about your brand or product,
the SDA includes a workflow (see Figure 9-3) to coordinate all of the steps
that are taken to analyze your social data.
Ingest
The first stage of social media analysis is ingesting social media messages. In
the SDA, this stage is handled by Streams, which has source operators for the
Gnip service (which provides messages from Twitter), and the BoardReader
service (which provides messages from microblogs—blogs, newsgroups,
forums, and so on). While Streams ingests the data into BigInsights, it
also does some initial filtering of spam messages to address issues of veracity.
Of course, it could be the case that you want to investigate this spam; per-
haps some of it isn't spam and you can evolve the filtration process through
customization—it's a learning environment, after all. Much like agile develop-
ment, the more times you iterate, the better the solution becomes. Streams also
provides real-time insights on the data as it's streaming into your organization.
Online flow: Data-in-motion analysis
Stream Computing and Analytics
Social Media
Data Sources
(Gnip, BoardReader)
Local Analysis
Global Analysis
Data Ingest
and Prep
Profile and
feedback
extraction
Visualization
Extended profile
entity analytics
Hadoop System and Analytics
Local Analysis
Global Analysis
Comprehensive
Social Media
Customer
Profiles
Define Sentiment
Targets
Profile and
feedback
extraction
Extended profile
entity analytics
Visualization
Offline flow: Data-at-rest analysis
Figure 9-3 The lifecycle of social data analysis
 
Search WWH ::




Custom Search