NLPeasy - a Workflow to Analyse, Enrich, and Explore Textual Data

Use pre-trained NLP-models, ingest into Elastic Search, and enjoy auto-generated Kibana dashboards!

Philipp Thomann

Data Science Elastic Search Machine-Learning Natural Language Processing

See in schedule Download/View Slides

Ever wanted to try out NLP methods but it felt it too cumbersome to set up a workflow for textual data? How to enrich your data based on textual features and explore the results?

NLPeasy ( does that: Enrich the data using well-known pre-trained models (Word embeddings, Sentiment Analysics, POS, Dependency Parsing). Then start the Elastic Stack on your Docker. Set-up indices and ingest it in bulk. And finally generate Kibana dashboards to explore the results.

Complicated? Not at all! Just do it in a simple Jupyter Notebook.

In this presentation we will give an architecture overview of the different components and demonstrate the capabilities of this Python package.

Type: Talk (30 mins); Python level: Intermediate; Domain level: Intermediate

Philipp Thomann


Philipp received his PhD in Mathematics from the University of Zurich in 2013. Afterwards he served as a postdoctoral researcher at the University of Stuttgart, Germany. As part of his research he worked on scalable machine learning packages and built liquidSVM, a Fast and Versatile SVM Package. His areas of expertise are machine learning, statistics, and data mining. Philipp joined D ONE team in 2017.