classification
ML-based text classification using a pre-trained model (e.g., fastText).
Options
| Option | Type | Default | Description |
|---|---|---|---|
MODELLOCATION | string | required | Path to the model file |
TOPK | integer | 1 | Number of top categories to return |
THRESHOLD | double | 0.0 | Minimum confidence score (0.0–1.0) |
Examples
CREATE TEXT SEARCH DICTIONARY classifier (
TEMPLATE = 'classification',
MODELLOCATION = '/models/cooking.bin',
TOPK = 3,
THRESHOLD = 0.5
);