Voxpopme integrates IBM Watson’s machine learning and natural language processing capabilities to deliver advanced video sentiment analysis.
The new sentiment analytics is powered by IBM Watson, which uses machine learning and natural language processing to identify the underlying sentiment within each individual sentence. This is used to process the transcribed text of any video in the Voxpopme platform, returning a polarity on every sentence of each video, determining whether it is positive, negative or neutral, with an associated score.
IBM’s system aggregates huge volumes of text data from social platforms to build an understanding of sentiment without the human biases that are often present in manual analysis. The shift towards automated sentiment analytics removes the subjectivity of human conclusions, vastly increasing speed, scalability and accuracy.
A deeper level of analysis is also available when using this with sentiment applied to Voxpopme’s Theme Explorer. Here, Theme Explorer provides a quick look view to demonstrate the sentiment breakdown of each theme identified within a video project to identify the most positive and negative sentences related to that theme.
Dave Carruthers, Voxpopme CEO commented:
“We’re delighted to be releasing our new sentiment analysis tools. Clients will now be able to understand the subtle nuances of every comment made.
We’re focused on accuracy at Voxpopme, so any new automation is assessed to discover the best blend of human and machine analysis. With sentiment, it’s clear that automation removes the margin for human error and increases speed and accuracy. In contrast, we’ll still be crowd-sourcing human transcriptions within 15 minutes as this still deliver far greater accuracy than machine transcription.
Along with our other platform features, sentiment analytics builds a deeper understanding of large volumes video, in less time, to establish tangible insights that can drive business outcomes.”