There’s a new feature in Watson Analytics for Social Media, the solution designed to make just about anyone socially savvy: you can view the actual mentions that are in the social data. While topic suggestions indicates what is in your data and provides context with word clusters, the new mentions view enables you to see the mention captured and analyzed by Watson Analytics for Social Media. For a more visual perspective of this feature have a look at this video.
At first glance, the visualizations for your Watson Analytics for Social Media appears to be similar to what you saw last month. But, when you elect to view the analysis, now you see an icon on the top right. This icon toggles an expandable view.
By expanding this view and selecting a bar, cell, point or area on the main chart in the analysis, you will see mentions that match the selection in the mention view.
Previously when you looked at the Watson Analytics for Social Media analysis, you may have asked, “Gee, I wonder what that spike of conversation is talking about in the chart for share of voice?” This feature will provide a sampling of mentions that enables you to peek into the content and get a better understanding of the spike.
You can scroll through the mentions or snippets to identify the patterns in the mention list. By understanding these patterns you can explain what is happening in the topics and themes in your analysis—which is key to your ability to deliver insight.
The left edge of each mention is color coded based on sentiment. Green is for positive, red for negative, grey for neutral and yellow for ambivalent. When the cursor hovers over the mention cell, you will see the evidence text underlined for the sentiment so you can identify why the mention is positive or negative.
In the sentiment chart you can see the distributions of sentiment. Now with the mentions view, you will be able see what conversations make up the positive or negative sentiment distributions. So you will be able to get a sense as to why a topic is so positive or negative. Again this information is only a click away.
Each mention cell includes the link to the original document so you can link back to where it occurred for additional context. Note that Watson Analytics for Social Media breaks down documents into atomic mentions which give much more detailed perspective on a topic or theme. If a blog post mentions one of your topics three times, and each with a different theme, you can capture each of the sentiments on the themes as they are typically separate mentions.
If you want to see what conversations are happening with top sites, sources and top authors, then this feature allows you to do so. When using the mentions view feature you may want to select the languages that you can interpret so you can understand the mentions that you are seeing.
Having mention view in theme breakdown visualization can help validate your data. You will be able to identify noise or missing definitions from the mentions view. From there you can iterate and develop the configuration to improve the quality of the analysis and resulting in a higher confidence level in the data. This confidence level introduces an important element in social media analytics. The ability to show what the data is made up of delivers credibility in the analytics which is crucial for your success in using social data as part of your overall analytics processes.
Did I mention, “If you are not already using this feature. You should try it out.”? Well now I guess I did 🙂 For a more visual perspective of this feature have a look at this video.