This app implements Role-Filler Binding (RFB) on text using a combination of NER and rule-based string parsing.
Roles and Fillers are analogous to key-value pairs, where the key (role) may correspond to a job title (e.g. Executive Producer), and the value (filler) corresponds to the named entity mention(s) filling that position. Binding applies an explicit relation between the two.
The current release of RFB uses a fine-tuned Bert-Base-Cased model to classify role/filler tokens and employs heuristic methods to determine relations between them. The intended target is OCR text extracted from visual media sources using upstream CLAMS apps, and the output is formatted as a raw CSV string.
General user instructions for CLAMS apps is available at CLAMS Apps documentation.
Below is a list of additional information specific to this app.
clams-python, mmif-python and their dependencies installed, to run the app locally.transformers for model inference.curl) to invoke and execute analysis.Run pip install requirements.txt to install app requirements.
docker to run the app in a Docker container (as an HTTP server).For the full list of parameters, please refer to the app metadata from CLAMS App Directory or metadata.py file in this repository.