At Google Cloud Next London, the company announced two new product updates. Its machine learning-based text meaning extractor API now supports more languages and has improved sentiment analysis. Additionally, its database service will soon enter into general availability.
The Cloud Natural Language API entered beta last July and Google has been incorporating customer feedback. With companies like Evernote and Ocado using it to analyze customer sentiment and derive action items from meeting notes, Google has expanded the API’s language support and improve the quality of the base natural language understanding technology.
Cloud Natural Language now supports entity, document sentiment, and syntax analysis in Simplified and Traditional Chinese, French, German, Italian, Korean and Portuguese, in addition to the existing English, Spanish, and Japanese.
The sentiment analysis can now understand specific blocks of text and works beyond just individual sentences or whole documents. Additionally, the quality has been improved with updated base models.
For example, for customer feedback such as “the phone is expensive but has great battery life,” users can now parse that the sentiment for phone is negative while the sentiment for battery life is positive.
The second update has Google’s globally-distributed database solution for third-party developers enter general availability on May 14th. During the three-month beta, Cloud Spanner has already seen a variety of uses:
- Powering mission-critical applications like customer authentication and provisioning for multi-national businesses
- Building consistent systems for business transactions and inventory management in the financials services and retail industries
- Supporting incredibly high-volume systems that need low-latency and high-throughput in the advertising and media industries