Don 8d896f1445 updated example properties 7 months ago
README.md Update 'README.md' 7 months ago
get_profile.js initial commit 7 months ago
package.json updated package.json 7 months ago
set_profile.js initial commit 7 months ago
template.app.properties updated example properties 7 months ago

README.md

datum-profiler

What is this project about?

Simple Node.js script to:

  1. Get FullContact profile from a Twitter handle.
  2. Get recent tweets of the user.
  3. Pass tweets to Watson Personality Insights to get buying preferences.
  4. Store enriched data in Datum network.

Use cases

  1. Personalized products and services.
  2. Customer profiling.
  3. Customer data validation.

How to run it?

  1. Get app usernames, keys and passwords from FullContact, Twitter and Bluemix services.
  2. Copy template.app.properties to app.properties file. Update app.properties with usernames, keys and passwords of third-party services above.
  3. Run "npm install" to install third-party modules.
  4. Run set_profile script with Twitter handle to store Watson-enriched profile in Datum.
  5. Run get_profile script with Twitter handle to retrieve above profile from Datum.

Run set_profile with a Twitter handle.

$ node set_profile.js kimkardashian
Person request resolved
Twitter auth resolved
Twitter request resolved
Watson request resolved
0x27a5ec7d2a423a1e41d522ef0d405beff88584b3c7b5789871f170d14ad7ff93
Datum store complete for kimkardashian

Then run get_profile to get the person's buying preferences.

$ node get_profile.js kimkardashian
Name
----------------------------------------------------------------------------
Kim Kardashian West
----------------------------------------------------------------------------

Websites
----------------------------------------------------------------------------
http://KKWBEAUTY.COM
----------------------------------------------------------------------------

Buying Preferences
----------------------------------------------------------------------------
Likely to prefer style when buying clothes
Likely to be influenced by brand name when making product purchases
Likely to be influenced by online ads when making product purchases
Likely to be influenced by social media when making product purchases
Likely to be influenced by family when making product purchases
Likely to eat out frequently
Likely to have a gym membership
Likely to like outdoor activities
Likely to like romance movies
Likely to like horror movies
Likely to like war movies
Likely to like drama movies
Likely to like rap music
Likely to like country music
Likely to like R&B music
Likely to like hip hop music
Likely to read entertainment magazines
Likely to read financial investment books
----------------------------------------------------------------------------

Try out other Twitter profiles.

node set_profile.js warrenbuffett
node get_profile.js warrenbuffett

node set_profile.js realdonaldtrump
node get_profile.js realdonaldtrump

node set_profile.js stephencurry30
node get_profile.js stephencurry30