Source: The Letters of Robert Browning and Elizabeth Barrett Barrett, Vol. 1 (of 2) via Project Gutenberg

Methodology: Scroll past the three visualizations below.


Map visualization of mentioned locations (circle radius proportional to frequency of mentions)

Locations in The Letters of Robert Browning and Elizabeth Barrett Barrett, Vol. 1 (of 2)
Locations in The Letters of Robert Browning and Elizabeth Barrett Barrett, Vol. 1 (of 2)

Wordcloud visualization of mentioned locations

Wordcloud of locations in The Letters of Robert Browning and Elizabeth Barrett Barrett, Vol. 1 (of 2)
Wordcloud of locations in The Letters of Robert Browning and Elizabeth Barrett Barrett, Vol. 1 (of 2)

Column visualization of top 10 mentioned locations

Column chart of locations in The Letters of Robert Browning and Elizabeth Barrett Barrett, Vol. 1 (of 2)
Column chart of locations in The Letters of Robert Browning and Elizabeth Barrett Barrett, Vol. 1 (of 2)

This is the sample output of a Python 3 Jupyter Notebook I created (publicly available via GitHub)

Read Me:

Jupyter Notebook (Python 3) guiding researchers through identifying location entities within text.

Specify the name/path of a text file. This notebook will use spaCy’s NLP entity recognition to automatically identify locations. Outputs include in-line text with recognized entities highlighted, a map visualization of proportional circles with the radius of each circle representing the frequency of the location in the text, a word cloud of the recognized locations, and a column chart of the top locations.

Make sure the text file is saved as ASCII or ANSI as Unicode may contain characters causing spaCy’s NLP processor to crash.

The NLP Entity Recognition using spaCy (https://spacy.io/) is based on this Towards Data Science blog post: https://towardsdatascience.com/named-entity-recognition-with-nltk-and-spacy-8c4a7d88e7da

Resulting interactive map visualization uses the Bing Maps geocoding API (http://www.bingmapsportal.com/) via the Geocoder library (https://geocoder.readthedocs.io/index.html). It is necessary to register for a freely available BING Maps Key to create the map visualization.

Resulting wordcloud visualization uses amueller’s word cloud generator (https://github.com/amueller/word_cloud).

Resulting bar chart visualization uses the standard matplotlib library (https://matplotlib.org/).

Visualizing Locations Mentioned in The Browning’s Letters

Leave a Reply

Your email address will not be published.