Once you've connected your data sources to Stratifyd, you can begin to analyze it and uncover insights. The model you apply depends on the type of data you have, how much data you have, and what questions you're trying to answer. 

Learn more in the following video, or scroll down for further details.

Sample use cases and the ideal model for each

When does each model work best?

Taxonomy - Best used when we have an idea of topics we want to track. When we categorize our unstructured textual data into thematic topics using a taxonomy, we create a new way to look at our data. We can compare the volume of data in each topic to see which one is more prevalent and trending over time. The unique thing about taxonomies, compared to other models in our platform, is that the model is entirely defined by the user. Unlike the machine learning models, there is no magic happening behind the scenes, so every result ties back to a specific rule that the user manually defined.

Theme Detection - Best used to reveal unknown topics in datasets for which we want to know the most common trends & themes. When we are unsure what topics are being discussed in our data, Theme Detection will bring to light our most prevalent conversations. Once you have ran a taxonomy to categorize topics you know exist, Theme Detection will reveal subtopics to dive even deeper into the category.

Theme Summarization - Best used to reveal unknown topics in datasets for which we want a more holistic view of the trends & themes. Once we're down to a set of data we care about (geo location, product line, taxonomy category, etc), Theme Summarization will pull out unique phrases, rather than the most common ones. 

Emerging Theme Analysis - Best used when expecting new conversations to appear (product launch, new survey questions, offerings change). This analysis will let us know any new conversations emerging in the data. We can keep a pulse on products and feedback to ensure we are are hyper aware of new topics.

Neural Sentiment Detection - Best used when we want to know how positively or negatively a conversation has become. We can pair this analysis with taxonomy and metadata to check in on positive and negative reactions.

Still have questions?

We're here to help! Don't hesitate to contact us for further assistance via chat or submit a ticket!