# IDF

### Description
Computes IDF of `T` in `DT` pairs
Strings are expected to be tokens.

### Input
- `DT [OBJ,STRING]`: a list of object-string pairs.
### Output
- `Tidf [STRING]`: IDF-weighted strings
- `DTidf [STRING]`: IDF-weighted pairs

### Parameters
- `Property Q`: The lower the `Q`, the more "obscure terms" are being penalised compared to standard IDF.
The range is `]0, 1]`. `Q=1` yields standard IDF. A suggested starting point s 0.6.

Output scores can be [aggregated](docs://score_aggregation) and/or [normalised](docs://score_normalisation).