Language support
IBM Watson™ Discovery offers two levels of language support, Basic and Full.
| Language | Support Level |
|---|---|
| English (en) | Full |
| Arabic (ar) | Basic |
| Chinese, simplified (zh-CN) | Basic |
| Dutch (nl) | Basic |
| French (fr) | Full |
| German (de) | Full |
| Italian (it) | Full |
| Japanese (ja) | Full |
| Korean (ko) | Full |
| Portuguese, Brazilian (pt-br) | Full |
| Spanish (es) | Full |
Basic support
- Discovery support includes:
- Document conversion
- Orchestration
- Discovery Tooling functionality
- Front-end API (excluding Front-end Query API)
- Language-optimized index
- Relevancy training
- Query expansion
- Stopwords
- Documentation in 9 languages: English, French, German, Japanese, Korean, Brazilian Portuguese, Spanish, Chinese Simplified, and Chinese Traditional. Scroll to the bottom of the IBM Cloud documentation to select the language.
- Enrichments
- Integration with IBM Watson™ Knowledge Studio custom models
- Language detection
To create a collection in Dutch or Chinese, you must use the API. For more information, see the API reference.
Full support
Full support includes basic support, plus:
- Passage retrieval
- Translated IBM Watson™ Knowledge Studio interface
- All Natural Language Understanding (NLU) enrichments, plus metadata extraction. NLU enrichments are:
- Entity Extraction
- Keyword Extraction
- Category Classification
- Concept Tagging
- Semantic Role Extraction
- Sentiment Analysis
- Emotion Analysis (English only)
English-only support
The following features are currently supported in English only:
- Element Classification enrichment
- Document deduplication (beta)
The Element Classification enrichment is deprecated and will no longer be available, effective 10 July 2020.
Japanese-only support
The following feature is currently supported in Japanese only:
- Custom tokenization dictionaries