Summarization of multiple documents into one sentence | |
Sentiment analysis (positive, negative, neutral, mixed) | |
Multi-language word association (amazing = asombroso) | |
Validated out-of-box sentiment precision (accuracy) | 93% |
Validated out-of-box sentiment recall | 97% |
Theme extraction (purchase intent, cost etc.) | |
Sentiment scored theme extraction (ex: negative cost) | |
Language detection (68 Languages) | |
Topic extraction and phrase detection | |
Topic clustering and relation | |
HTML handling (including broken HTML) | |
Extensive unicode support, including malformed data |
Throughput | Up to 1 Gbps (upgrade options available) |
Processing Latency | ~1 millisecond |
Supported Character Formats | UTF-8 and HTML |
Returned API Data Format | JSON (XML optionally available) |
Maximum Dataset Size | 16,384 petabytes |
Supported Languages | Detection of 68 languages, sentiment for English, French, German, Spanish, Portuguese, Financial English |