Stratifyd’s Emerging Theme Analysis feature helps you identify “fresh topics”—emerging trends, themes, and keywords/phrases—within a specific time comparison in your data set. Emerging Theme Analysis automates the process of data monitoring in order to save you the time and energy required to do so manually. When paired with other Stratifyd features, specifically Listen Actions, Emerging Theme Analysis takes on the role of a virtual 24/7 analysis assistant.
Best Practices for Emerging Theme Analysis
While automating the task of continuous data monitoring within the platform provides for easier analysis, there are several practices that we suggest in order to mitigate the effects of the feature’s limitations.
1. Use the Emerging Theme Analysis feature when trying to identify topics that could be new and/or anomalous.
Let’s imagine I’m analyzing mobile app reviews. Then, imagine I’m tracking for positive and negative review behaviors, but not for bugs. Using the Emerging Theme Analysis feature, I can identify a spike in my app failing to work on specific devices. In response, I can add a label for app bugs/malfunctions.
2. Use data with date ranges where new topics can be discovered.
Emerging Theme Analysis can detect and compare based on the following cadences:
NOTE: Emerging Theme Analysis is best suited for non-static streams. The frequency with which the data set updates should depend on data volume.
3. When beneficial, edit the advanced settings to modify filters to allow for comparison analyses based on one specific time range.
Let’s imagine your company has launched a new product. In a situation like this, you might modify filters to compare reviews for the new product to reviews for existing products in order to detect new topics being discussed.
NOTE: Rolling date filters will be the default settings (e.g. a rolling weekly cadence will compare the past 7 days to the previous 7 days).
In its initial version, the Emerging Theme Analysis feature does have a few limitations. A brief description of each is provided below.
Based on research with the Emerging Theme Analysis feature, an average of roughly 200 records are required in a data set in order for 2 to 3 meaningful new topics to be generated. So, if your data set receives less than 200 new records per week, a weekly cadence won’t produce results. While this is not inherently an issue, the feature currently includes no in-widget messaging that communicates that there are no new results. Instead, it’s important to recognize that some users may misinterpret this lack of information as a functionality error or malfunction.
Topics with a fresh value less than 0.5 are not really “fresh” or emerging given that they were likely included in the previous comparison data range. It’s important to recognize that these “fresh values” (values less than 0.4) can be hidden from display in the widget in order to prevent unnecessary user notifications by the accompanying Emerging Theme Analysis feature.
How to Use the Emerging Theme Analysis
1. You can create a model from within a workspace, or you can add one to the Models tab. Here, we create it from within a workspace by selecting the workspace tab in the navigation pane.
2. To access the Data Settings Panel menu, click the Data settings button, accompanied by the gear icon.
In the data settings panel that appears, you'll see a list of all data connections for that workspace. Make sure you have selected the data stream you want to work with in the Connected column.
3. In the Analyze tab, expand the section labelled What are our customers saying? to see available models. Choose Emerging Theme Analysis by clicking the + icon.
4. For the model to run successfully, you'll need to make three selections: comment text, time stamp, and a rolling date interval.
A note on rolling dates:
5. Depending on the size and complexity of your data, it may take some time for the analysis to finish running. When you return to the Data Settings, you'll see this analysis within the Deployed section at the top of the the Analyze tab.
We're here to help! Don't hesitate to contact us for further assistance via chat or submit a ticket!