Textplorer

Textplorer is a web-based visual analytics tool to help the qualitative analyst in the Grounded Theory Method. Grounded Theory is a general method of developing theoretical constructs inductively from data sources gathered as a part of a qualitative research study. It uses a "construct-oriented approach" where the analyst creates categories based on patterns in the data. These categories are further grouped into higher-level categories, and relationships (especially causal) between categories are identified. Finally a "core concept" is chosen among these categories, and based on the relationships identified, a theory is formed.

Textplorer uses natural language processing (NLP) techniques along with interactive text visualizations to help the user explore the text data, identify concepts of interest, and establish relationships between these concepts.

Publication

Textplorer was developed as part of a paper presented at EuroVis 2017. The full citation, link to the pdf, and a video demo are shown below.

Chandrasegaran, S., Badam, S.K., Kisselburgh, L., Elmqvist, N., and Ramani, K. Integrating Visual Analytics Support for Grounded Theory Practice in Qualitative Text Analysis. Computer Graphics Forum (Proc. EuroVis), 36 (3), pp. 201–212, 2017.

(Download Paper)

Installation Instructions

We use the Stanford Parts-of-Speech (POS) Tagger and the Stanford Named Entity Recognition (NER) Tagger, which require a little setup in Java. So before installing Textplorer, perform the following steps first.

Install Dependences

1. Install JDK (v > 1.8), and add the path to the bin file (e.g. C:/Program Files/Java/jdk1.8.0_101/bin or /usr/bin/java to the environment variable JAVAHOME (for Windows), or to your $PATH variable (for Linux/OSX).
2. Download the Stanford Named Entity Recognizer and extract its contents to a folder, say $HOME/StanfordNER/
3. Add $HOME/StanfordNER to the CLASSPATH environment variable (for windows) or to your $PATH variable if linux/OSX.
4. Download the Stanford CoreNLP models and extract the files. Copy the folder nlp from stanford-english-corenlp-20YY-MM-DD/edu/stanford/ and place it in $HOME/StanfordNER.
5. Download the Stanford POS Tagger, extract the contents into a folder, say $HOME/StanfordPOS. Add this folder to your $PATH variable.
6. Make sure Node.js and Python are installed on your system
7. Install the NLTK package for Python.
8. Run python and the command line and type import nltk, followed by nltk.download(). This opens a download window titled "nltk downloader". Select the "Corpora" tab, and download the "wordnet_ic", and "stopwords" corpora.

Install Textplorer

Textplrorer's server uses the Node.js runtime, so it needs to be installed first, in order to install all related libraries.

1. Download the source code as a zip file or clone the repository by typing the following command on the terminal:

   $ git clone https://github.com/senthilchandrasegaran/textplorer.git

2. On the Node.js Command Prompt (for Windows) or Terminal (for OS X), navigate to the main folder (the one that contains the file textplorer.js) and type in:

   $ npm install

This installs all required modules.

NOTE: It is wise to update the path variables in the file public/pythonscripts/infoContent.py to reflect the paths of the Stanford NER and POS Taggers, instead of relying on the $PATH variable setting. Make sure you have the absolute path (paths starting with ~/ or $HOME will not work.)

License

Released under BSD license. Copyright 2017 Senthil Chandrasegaran.