Follow Datanami:
March 16, 2018

More Emojis Means More Consumer Data

Via Shutterstock

Emojis are being re-enlisted in the never-ending struggle among marketers to divine consumer preferences.

In response to the release of a set of 157 new digital facial expressions used among other things to convey consumer attitudes, machine learning and AI specialist Lexalytics said this week it would support the new batch of emojis approved by the Unicode Consortium.

The vendor’s AI-based text analysis platform can be used to discern consumer sentiment when the new emojis are used across social media, email or other forms of text communications. “In order to understand the full scope of what people are saying and feeling about a particular product, brand or service, emojis are becoming just as critical as standard text for our enterprise customers,” said Lexalytics CEO Jeff Catlin.

The Boston-based analytics company pitches its “words first” AI and machine learning platform as a hybrid approach to text analytics. While conventional systems are based solely on machine learning techniques requiring historical data to train the system, Lexalytics’ platform provides real-time “direct tuning” that can be implemented in advance. The approach is said to save many hours of training, time that can be better used to understand and translate the new emojis.

The company’s AI-based “natural language understanding” platform combines machine learning with dictionaries and natural language processing code. The approach is distinguished from frameworks such as supervised machine learning using neural networks.

“We can both ‘tune’ and ‘train’ our system,” Lexalytics noted in a blog post. “Tuning means just reaching in and placing a line in a file (or turning a knob) that tells the system exactly what we want it to do. Training means gathering a set of examples, annotating them if necessary, and teaching the system through these examples.”

Hence, Lexalytics claims to be the first text mining vendor to calculate emoji sentiment value along with the semantic value contained therein.

A version of the Lexalytics text analyzer released in October 2016 expanded its machine learning capabilities along with the ability to ingest email databases. The platform is geared to social media marketers and “customer experience” managers sifting through customer emails as well as text related to customer reviews that include the growing number of emojis.

The company’s emphasis has steadily shifted toward gleaning insights from the consumer data expected to be embedded the latest batch of emojis.

The new set approved by the Unicode Consortium will include data needed by vendors to begin integrating the code into their platforms ahead of the launch of Emoji 11.0 scheduled for June. The new emojis are expected to show up on mobile phones in August or September.

The set of 157 symbols brings the total number of emojis to 2,823.

Recent items:

Using Emojis to Boost Sentiment Analysis

The Data Science Behind Dollar Shave Club