Natural Language Models

Natural Language Models

Natural Language Models

Natural Language Models

TAG: METHOD

970af46b50aece2e422fdf2ed4ef33df

When to Use It: 

You have a large volume of freeform text that you would like to mine for insights — e.g. which keywords, or named-entities are referenced most frequently? What is the customer’s sentiment? What metaphors do customers use when describing your brand?

What It Tells You: 

  • Which keywords or named-entities are referenced most frequently
  • Which topics are most commonly discussed
  • The metaphors customers use to describe your brand

 

What Kind of Data it Needs / Takes:

  • Statements / comments on a discussion forum
  • Articles, blog posts
  • Open-ended survey questions

 

Statistical Methods: 

Advanced NLP models, including:

  • Keyword extraction (e.g. RAKE, TextRank)
  • Topic modeling (Latent Dirichlet Allocation)
  • Metaphor recognition

When to Use It: 

You have a large volume of freeform text that you would like to mine for insights — e.g. which keywords, or named-entities are referenced most frequently? What is the customer’s sentiment? What metaphors do customers use when describing your brand?

What It Tells You: 

  • Which keywords or named-entities are referenced most frequently
  • Which topics are most commonly discussed
  • The metaphors customers use to describe your brand

 

What Kind of Data it Needs / Takes:

  • Statements / comments on a discussion forum
  • Articles, blog posts
  • Open-ended survey questions

 

Statistical Methods: 

Advanced NLP models, including:

  • Keyword extraction (e.g. RAKE, TextRank)
  • Topic modeling (Latent Dirichlet Allocation)
  • Metaphor recognition

When to Use It: 

You have a large volume of freeform text that you would like to mine for insights — e.g. which keywords, or named-entities are referenced most frequently? What is the customer’s sentiment? What metaphors do customers use when describing your brand?

What It Tells You: 

  • Which keywords or named-entities are referenced most frequently
  • Which topics are most commonly discussed
  • The metaphors customers use to describe your brand

 

What Kind of Data it Needs / Takes:

  • Statements / comments on a discussion forum
  • Articles, blog posts
  • Open-ended survey questions

 

Statistical Methods: 

Advanced NLP models, including:

  • Keyword extraction (e.g. RAKE, TextRank)
  • Topic modeling (Latent Dirichlet Allocation)
  • Metaphor recognition

When to Use It: 

You have a large volume of freeform text that you would like to mine for insights — e.g. which keywords, or named-entities are referenced most frequently? What is the customer’s sentiment? What metaphors do customers use when describing your brand?

What It Tells You: 

  • Which keywords or named-entities are referenced most frequently
  • Which topics are most commonly discussed
  • The metaphors customers use to describe your brand

 

What Kind of Data it Needs / Takes:

  • Statements / comments on a discussion forum
  • Articles, blog posts
  • Open-ended survey questions

 

Statistical Methods: 

Advanced NLP models, including:

  • Keyword extraction (e.g. RAKE, TextRank)
  • Topic modeling (Latent Dirichlet Allocation)
  • Metaphor recognition

USE CASES

a21861f8f3edc077cac7a616ee8dce55

Refreshing A Global Employer’s Value Proposition To Recruits

Refreshing A Global Employer’s Value Proposition To Recruits

ABOUT

We are marketers and technologists that build robust models to guide managers in their decisions. We believe in sound design, interpretability, and going out and getting the data if you don’t already have it.

ABOUT

We are marketers and technologists that build robust models to guide managers in their decisions. We believe in sound design, interpretability, and going out and getting the data if you don’t already have it.

ABOUT

We are marketers and technologists that build robust models to guide managers in their decisions. We believe in sound design, interpretability, and going out and getting the data if you don’t already have it.

ABOUT

We are marketers and technologists that build robust models to guide managers in their decisions. We believe in sound design, interpretability, and going out and getting the data if you don’t already have it.

ABOUT

We are marketers and technologists that build robust models to guide managers in their decisions. We believe in sound design, interpretability, and going out and getting the data if you don’t already have it.

Join Our Community

Join Our Community

Join Our Community

Join Our Community

COPYRIGHT GRADIENT METRICS 2018

COPYRIGHT GRADIENT METRICS 2018