Natural Language Processing Package

The Natural Language Processing (NLP) package enables bot builders to analyze, evaluate, and take action based on the processing of text.

The source code for this package is available on the Automation Anywhere GitHub page

Top Benefits

Enable bots to better understand free form text
Analyze public customer reviews, tweets, and mentions for positive/negative sentiment
Evaluate text of incoming emails or chat messages to detect parts of speech
All processing occurs locally using a pre-trained Stanford NLP library

Actions

Sentiment Analysis enables the evaluation of text for positive, negative, and neutral sentiment
Entity Extraction enables the identification of people, locations, organizations, etc.
Part of Speech Detection enables extraction of text based on the part of speech each word represents.

The Natural Language processing package enables bot builders to build intelligence into their automation through the analysis of free form text. This package is designed to perform Sentiment Analysis, Named Entity Recognition, and Part of Speech detection to provide Automation Anywhere bot builders with the ability to analyze text against a pre-trained NLP library provided by Stanford NLP Core.

Best of all, all NLP processing occurs on your local system - no external API calls/subscriptions required. All processing occurs locally.

Note: Because this package includes the entire Stanford NLP Core library, it's larger than most other Bot Store packages. Upon first running the included demo bot, the bot runner will download the entire package for processing (~450MB). This download will only occur once. On subsequent bot executions, that package would have been cached locally and will be used for any follow bot runs that reference this package.