text annotation uses on WebChartAI
extracting actionable insights to drive better decisions

text classification

Text classification enables natural language processing to sift through

massive datasets to extract specific information needed to meet an objective.  

Each of the use cases described on this page are forms of text classification, and vary based on the project's objective. 

 

Entities of all kinds - commercial, governmental, non-profit and others - use text classification to automate tasks like tracking brands and consumer preferences; improving customer support services; moderating content; improving search results; and tracking public opinion, to name a few. 

 
 

Intent analysis trains a natural language processing engine to look for clues within text or speech that signal a desire or objective.  Once an intent is understood, it can be classified in order to apply an appropriate response. 

​

WebChartAI is perfect for creating datasets that teach natural language processing applications to accurately identify intent.  

intent analysis

automation that accurately

interprets desires or objectives

dictionaries

& word lists

Classify words and meanings in order to  normalize usage in automation applications 

A word list is a collection of words relevant to a topic.  For example, a  security agency might build a word list of terms used by threat agents that refer to planned operations for monitoring purposes.  

​

A dictionary is a collection of meanings associated with a word. For example, an NLP application trained to monitor for illegal drugs would include bliss in its dictionary as a meaning for Fentanyl. 

 
 

intent variation & utterance collection

Intent variation is understanding what someone wants, while utterance collection is understanding all the different ways someone may ask for it. 

Intent variation defines all possible meanings of a query.  For example, "warranty question" could mean buying

a new  warranty, entering a claim on an existing warranty, or clarifying a warranty's terms, to name a few. 
 

Utterance collection is defining all the ways intent is expressed,  such as "I'd like to buy a warranty" vs. "how do I purchase a new warranty?".

Both utterance collection and intent variation play an important role in enabling natural language processing to add automation to tasks like chatbot responses, customer service interactions and other business processes. 

Over 15 million minutes of audio and 50 million documents

have been processed using WebChartAI's workflow tools. 

sentiment analysis

accurately interpreting attitudes, emotions and opinions

Sentiment analysis helps natural language processing accurately interpret a target audience's attitudes, emotions and opinions. 

​

Accessibility to large collections of data via social media monitoring and data mining makes sentiment analysis one of the most widely-used - and important - classification techniques available. 

 

content moderation & spam detection

interpret and classify content for appropriateness and spam detection

Content moderation and spam detection applications use natural language processing to classify objectionable text-based content.  

​

WebChartAI is used to create training datasets for content moderation and spam detection automation.  

 

text summarization

summarize text based on situational

needs, reader types or access rights

Text summarization uses natural language processing to identify key words, phrases and sentences within a body of text and preserves them within a condensed version of the original.   

​

WebChartAI's workflow tools are used to manually summarize bodies of text in order to generate training data for model building and improvement.  

language detection

identify languages in speech and text-based

interactions for workflow-based automation

Language detection enables natural language processing to add automation to workflows where multiple languages are used. 

 

It's often used as a first layer of data  processing prior to the use of other  techniques, such as sentiment analysis.

 
 
 

verbatim transcribing

highly accurate transcription with rapid

turn-around for custom model tuning

Over 40 million minutes of audio have been transcribed by over 20,000 transcriptionists in the last ten years on platforms developed and owned by WebChartAi - WebChartMD for health related transcribing (webchartmd.com) and Xelex Digital for non-health (xelexdigital.com).

​

WebChartAI's robust workflow was designed for high-volume verbatim transcribing.  Custom  model tuning projects of 500 - 1000 audio hours can be verbatim transcribed and returned with 48 hours if required.  

WebChartAI