Analysis of technology options for the categorisation and understanding of multilingual customer feedback.
For CDO’s, CXO’s, heads of product and customer service teams, understanding customer feedback is the most fundamental task to provide good customer service. However, there are two major obstacles to automatically detect the meanings of customer feedback in the global multilingual environments to international companies. First, there is no widely acknowledged categorisation (classes) of meanings for customer feedback. Second, the applicability of one meanings categorisation, if exists, to customer feedback in multiple languages is questionable.
As a company scales its customer base internationally it can become a challenge to monitor customer satisfaction and ensure accurate responses, ADAPT was challenged to analyse the performance of existing data processing pipelines and recommend areas for enhancement and propose technology options.
Adapt assessed the performance of Microsoft’s internal tools for customer feedback analysis in English, Spanish and Japanese.
The first corpus of customer feedback in English, French, Spanish, and Japanese annotated using a taxonomy of six high-level semantic labels, was developed.
A new five-class categorisation for customer feedback was suggested as a method of streamlining feedback.
This led to high levels of classification accuracy.
This solution eliminated the need to develop native-language classification technology.
This in turn leads to a comprehensive understanding of international customer feedback and a globally consistent brand voice. This will improve brand loyalty and trust.
Auto-categorise international feedback; measure satisfaction across languages; monitor quality across languages.
A #ClimateNeutral Ireland is the future, but how can we get there by 2050? 🌐
If your research touches on this problem then the newly launched €65 million SFI National Challenge Fund is looking for ambitious researchers who are ready for a challenge.