A startup sentiment analysis suggests that Twitter Inc (NYSE:TWTR) loved Apple’s WWDC keynote.
The startup sentiment analysis that goes by the name TheySay is the brainchild of scholars from Oxford University, and its development was headed by the professor of Computational Linguistics. The claim that Twitter loved Apple’s WWDC keynote came from one of the lead researchers from the University.
According to Dr. Karo Moilanen, TheySay uses a sentiment analysis algorithm to follow human reasoning by observing the news, social networks, and other sources. It keeps track of key phrases and keywords that are oriented towards specific topics. The keywords are then passed through a series of analysis that result in sentiment scores. The score may be negative, positive or neutral.
The Sentiment analysis also generates a speculation intent analysis and multi-dimensional emotion scores. The latter is based on the emotional responses detected. TheySay makes use of automated Natural Language Processing Deep Learning and Machine Learning methods to imitate human responses when they read text.
The Oxford startup made an analysis of 94,528 tweets starting half an hour before the event began and ended right after the peak of the event. The results of the analysis suggested that Apple is the company with the highest positive ratings.
The report also showed that Apple’s conference had a 93% positive sentiment during the whole period that the study was carried out. Keynote received a 94% regard from the analyzed tweets. Aside from the number of women in attendance being limited to two, who also happened to be white, the analysis from TheySay still suggested that Apple efforts to enhance diversity were well embraced.
The sentiment analysis also featured a few of Apple’s products and services. Apple Music received the lowest positive sentiment with a score of 85%. Apple pay received the best ratings with a positive sentiment score of 98%. A few other inbetweeners including OS X 10.11: El Capitan and iOS 9 received 88% and 98% ratings respectively.
