Based on the state-of-the-art deep learning and Natural Language Processing (NLP) research coming out of Silicon Valley, Onclusive’s Global Sentiment Analysis leverages artificial intelligence to deliver automated insights around the tone of earned media coverage. Sentiment scores are delivered in real-time through Onclusive’s media monitoring and analytics platform.
Why is sentiment analysis important for PR teams?
PR teams can no longer measure their effectiveness or the overall nature of company coverage simply based on the volume of articles they have secured within their target publications. Onclusive empowers PR teams to prove the broader value of their work.
We invented Power of Voice™ to measure the quality of communications content, and invented PR Attribution™ to measure its actual bottom-line impact. One of the best measures of quality, and a key factor within the Power of Voice™ metric, is the tone & sentiment of media coverage over time. Further, accurate sentiment understanding is critical when researching topics and subject matter related to companies’ communications strategies.
How is Onclusive’s AI-driven sentiment measurement engine different from other sentiment analysis solutions on the market?
There are two primary differentiators: Accuracy & Number of Languages.
First, we believe our proprietary model delivers the most accurate interpretation of the tone, feeling, and sentiment of the article. We have conducted numerous experiments to compare other third party solutions and have consistently demonstrated a statistically significant improvement in accuracy levels.
Second, the current third-party sentiment solutions in the industry support only a handful of languages. This creates challenges for businesses who are earning media coverage in multiple markets globally. The Onclusive multilingual model will support 104 languages — significantly more languages than any other provider that we are aware of.
How is sentiment measured?
Sentiment is demonstrated using a three-factor score: Positive, Neutral, Negative.
Sentiment can then be reported at three levels:
- Aggregate sentiment for your total media coverage during a specified date-range
- The sentiment of a specific article
- The sentiment of each entity within the article – entities are persons, places, things contained in the article and our model provides the sentiment for each entity in the context of the other entities and the overall article
Users of the tool can also manually adjust the sentiment score for an individual article based on their own personal interpretation.
When will this be available to Onclusive customers?
The first version has been released this month and we will be deploying enhancements and additional languages over the next few months.
“This new innovation represents the next evolution of AI-driven communications technology and our corporate commitment to advancing the state-of-the-art in media monitoring. I am thrilled to see our product continuing to advance in this way and am excited for the new possibilities that this brings to our industry.”
–Dan Beltramo, CEO at Onclusive