• Home
  • News
  • DigitalMR launches APIs for automated text and image analysis

DigitalMR launches APIs for automated text and image analysis

Consumer insights technology agency DigitalMR has developed application programming interfaces (APIs) for automated sentiment, emotions, topics, noise elimination (text) analysis, as well as image brand presence, text extraction and theme analysis.

Such APIs can be used by market research agencies and end-clients in any sector or industry, to analyse unstructured data from any source – for example open ended responses from surveys or transcripts from online communities or in depth interviews, social media posts, call centre transcripts etc., to identify whether it is relevant to them and the sentiment or emotions expressed, as well as topics of conversation. The image analytics APIs make it possible to detect a brand or product logo within images posted by consumers online, and to discover occasions where products are in use by consumers.

Sophia Papagregoriou, Sales & Marketing Manager of DigitalMR, said: “This launch enables DigitalMR to service sophisticated MR agencies/clients with an innovative deliverable (for market research standards) that is way cheaper and more flexible than dashboards or powerpoint reports.”

The newly released APIs are available as generic in the English, Spanish and Arabic languages, or specific to a product category (for a limited number of categories) in a wide range of languages including Traditional Chinese, Vietnamese, and Portuguese, and can be utilised to analyse any unstructured text, in a fast and effortless way.

DigitalMR owns proprietary solutions for market research, specialising in the use of Artificial Intelligence for customer insights, and has been developing custom machine learning models for automated sentiment and emotions analysis of social media posts and other unstructured data for several years, reaching high precision in numerous languages and for various subjects and product categories.

Web: https://www.digital-mr.com/