ConceptNetwork demo

A demo versions of a software package for the discovery of a network of related concepts from unstructured texts in the DARIAH-FI project.

Tutorial

Requirements

1. Jupyter Notebook installation with R kernel (via IRkernel R-package).

2. Your own data. For example, download FSD3360 Helsingin Sanomat Loneliness Survey 2014 from the Finnish Social Data Archive. Use file format which can be imported into R.

3. The input data for the tool must be in CONLL-U formatted CSV. See UDPipe package for R how to convert data to CONLL-U format.

Pipeline

Clone this repository and open it Jupyter notebook.

Order	Input	Notebook	Output
0.	Any data file imported into R	Data Preparation Tutorial.ipynb	csv file in CONLL-U format
1	csv file in CONLL-U format	Concept Exploration.ipynb

0. Data preparation

See Data Preparation Tutorial.ipynb notebook.

1. Concept Exploration

See Concept Exploration.ipynb notebook.

In Development

• Manual Concept Extraction.ipynb