Written language can be broken down in a logical way, from paragraphs to sentences, words to letters. But, it’s when we combine the meanings of each smaller piece that we can understand the entire message. We derive meaning from social posts by breaking them down in a similar way and extracting as much information as possible about a post’s meaning and author. When all the pieces come back together, we get a richer understanding of the context and interconnections that exist in consumer data from social media.
Getting to that true meaning, however, is messy. Searching through posts using Boolean queries can be disorganized, time-consuming, and non-intuitive. This is where our taxonomy of over 25,000 classifiers (everything from brands, celebrities and movies to emotions and author demographics) helps. By leveraging multiple techniques including machine learning, linguistics, and keyword search, we give structure to social data and make it easier for Kairos users to quickly find what they’re looking for, with minimal noise.
To make effective use of our 25,000 classifiers, we organize them into a logical taxonomy. Here, it’s helpful to imagine a hierarchy like a family tree. A broad classifier like “cinema” may be toward the top. From there, the category can be subdivided by genre (comedies). From there, we branch out to sub genres (romantic comedies), then specific movies (Trainwreck) and actors (Amy Schumer).
Both the individual classifiers and relationships among them are important. In some cases, specific interests provide the key to deriving insight. If, for instance, Trainwreck conversations skyrocketed, the full “Entertainment” category is too general to clearly pinpoint underlying trends. In other cases, higher-level categories can help uncover macro-level trends that any single interest doesn’t explain on its own.
Networked Insights’ classification taxonomy provides a diverse and robust lens through which marketers can understand their audiences. It’s akin to viewing a picture in the entire spectrum of vibrant colors versus seeing in only eight primary colors. Through our taxonomy, we’re able to harness a more complete depth of human knowledge and deliver more meaningful insights to brands. Contact us to learn more.