OMTD-SIMPLE-WORKFLOWS

Provides scripts and code for executing some text mining workflows within a docker container.

Installation

Step 1: Clone omtd-simple-workflows by typing

git clone <repoURL>

Step 2: CD to omtd-simple-workflows directory that has been created. For building the projects type

./installProjectsDepedencies.sh
mvn clean install

Step 3: Create a docker image (named omtd-simple-workflows-docker) that contains everything that is needed.

./omtd-simple-workflows-createDockerImg.sh

Step 4: Use the following commands to a) create a container from the image produced by the previous step b) start the container c) get a bash shell inside the container.

sudo docker create --name omtd-simple-workflows -t omtd-simple-workflows-docker
sudo docker start omtd-simple-workflows
sudo docker exec -i -t omtd-simple-workflows  /bin/bash

Examples

Inside the container try the following examples. The same examples can run and in the host machine.

Example 1:

# PDF2XMI example with DKPRo PdfReader. Reads a folder (../testInput) with PDFs and creates an output folder (../testOutput) 
# with the respective XMIs that were produced.
./Linux_runDKPro_PDF2XMI_example.sh
# Check that the produced output is correct.
./checkDiff.sh ../testOutput/ ../testOutputPDFToXMIRef/

Example 2:

# Chebi example from UNIMAN. Reads a folder (../testInput3) with PDFs and creates an output folder (../testOutput2) 
# with the respective XMIs that were produced
./Linux_runUNIMAN_Chebi_example.sh

Example 3:

# Topic Inference with DKPro (Mallet LDA). Reads a folder (../testInput) with PDFs and creates an output folder (../testOutput3) 
# with the respective XMIs that were produced
./Linux_runDKPro_PDFLDAInference_example.sh

Example 4:

# Named Entity Resolution with DKPro (Stanford NER mode) for Social Science and Humanities. Reads a folder (../testInput2) with PDFs and creates an output folder (../testOutput4) 
# with the respective XMIs that were produced
./Linux_runDKPro_PDFNERInference_example.sh

NOTES: All scripts are available at omtd-simple-workflows/scripts directory.

mandiayba / omtd-simple-workflows Goto Github PK