Data Services with SAP Lumira – Analyzing Tweets – Part 2
After my first blog where I share what I am tweeting about Using Data Services Text Data Processing with SAP Lumira to Analyze Tweets , I now want to take a look at who I am tweeting with using Data Services . Data Services can be used to read the 140 characters of unstructured data to read the twitter ID of my tweets. Who was I tweeting most with? Ingo Hilgefort ? Fred Verheul ? Would Klout tell me this?
Using Data Services as shown above, Social Media/Twitter ID is one of the entity types.
Run the Data Services batch job I created and send that to a file, to be consumed by SAP Lumira (formerly SAP Visual Intelligence)
Because there are so many people I tweet with, use the gear icon to rank the top people I tweet with.
Above shows the top 15 individuals in my tweets. Many are fellow ASUG volunteers, such as Karin Tillotson Susan Keohan
I have never met Fred Verheul or Tobias Trapp in person yet I feel I know them personally through twitter.
Now Lumira shows the usual heat map, shading those I’ve tweeted with by range.
The above shows the usual heat map. Because my top hashtag was #SCN in my previous post, it is no surprise that SAPCommNet is one of the top twitter handles in my tweets.
I gave the radar chart a try; not sure how helpful it is with this analysis.
I happen to like the tag cloud when doing text analysis.
The waterfall chart is visually interesting but doesn’t tell me anything new.
I wasn’t surprised at these results because I work frequently with Ingo and Ina Felsheim as both are our SAP points of contact to ASUG. Thorsten Franz is a wonderful conversationalist on twitter, along with Jon Reed .
Interestingly, out of the above people I tweet with, only Vijay Vijayasankar shows up as my Klout “influencers”.
Still it was fun to analyze who I am tweeting with. For sure companies should be looking at their own social media accounts, analyzing what and who they are tweeting with – and to me, Data Services makes this happen quite easily.
Is it good enough for the Data Geek contest? Also see Get Your Data Geek Badge Today! by Anita Yuen
This is the sort of analysis I wish I'd thought of doing 🙂 I know what I'll be doing next time I fire up Lumira...
Thanks for the idea, Tammy!
I'd be interested in your results, for sure 🙂
Hey Tammy,
Honored to be among the top 15 together with some great people and twitterati. I'm pretty sure you would also be in my top 10 🙂 .
And very nice visualizations. I had no idea Lumira could do this. Well, I probably should have suspected something like this, but still. Well done!
Cheers, Fred
Interesting results, Tammy.
You could also use the Text Data Processing feature of Data Services to graph the sentiment of all of the tweets that mention you, for example. Although we know they'd all be wildly positive. Is that a sentiment category? 🙂
Could you include an "eim" tag in the article so the EIM space picks it up?
Hi Ina,
Thank you for reading and commenting.
EIM tag has been added
I will look at including the other things you mention in a "how to" for Data Services
Tammy