hltcoe/lome repository overview

⁠LOME: Large Ontology Multilingual Extraction

The hltcoe/lome Docker image requires NVIDIA-Docker support.

The hltcoe/lome Docker image reads in an input directory of text files and writes Concrete⁠ Communication document files to an output directory.

The text files in the input directory are expected to have the format:

• all text files end with the extension .txt
• sentences are separated by newlines
• tokens are separated by whitespace

The hltcoe/lome Docker image supports both the AIDA and KAIROS entity typing ontologies, which can be specified with the --aida or --kairos flags. The default entity ontology is AIDA.

The hltcoe/lome container can be invoked using:

docker run --gpus all --rm hltcoe/lome [--kairos] <input_directory> <output_directory>

In order to read and write files on the Docker host, the container needs to be run with appropriate Docker volume⁠ flags (e.g. -v/--volume/--mount).

If the working directory on the Docker host contains a sub-directory named text-files/, then the container coud be run using:

docker run --gpus all --rm -v `pwd`:/mnt hltcoe/lome [--kairos] /mnt/text-files /mnt/comms

The container will treat all files with a .txt extension in the text-files/ subdirectory as text files. The container will create a sub-directory on the host named comms/ with annotated Concrete Communications.

⁠Disabling pipeline components

The time component of the LOME pipeline can be disabled with the --disable-time flag, e.g.:

docker run --gpus all --rm hltcoe/lome --disable-time <input_directory> <output_directory>