No description, website, or topics provided.
TypeScript HTML JavaScript CSS
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
e2e
tsapp
.gitignore
bower.json
index.html
karma-test-shim.js
karma.conf.js
package.json
protractor.config.js
readme.md
systemjs.config.js
tsconfig.json
tslint.json
typings.json

readme.md

Good or Bad Images

Angular2 frontend to do training on images.

♫ What is it good for?

(♫ absolutely nothing ♫) A friend asked to make a traning app that would help her train people. So why not practice doing it with her requirement.

Flow

Backend: Node with Postgres

Page 1: generic home.

At this level the user should already be loggued-in via Google & Passport (or else).

Page 1.1: Main Test

Algo: Randomize the selection on only img the user never saw. (this is done in Postgres directly) The user will have a 1 images to inspect at the time

  • List of Y criteria/defect to select from.
  • Multiple item can be selected. All required item need to be selected to succede the images. (Ranges??? Optional???)
  • TODO: Each images can have a Wiki/reference link to learn from.
  • The page will have a comment box at the bottom for the user to input comments.
  • When submit it will tell you if it was good or wrong.
  • TODO: If wrong a button will be there to contest the default img value (reason), or simply click "next". Data To Save:
  • Time between open and Submit
  • Good Fail or Failed Good img
  • What was missing in the defect
  • What was extra in the defect
  • Comments on the img.
  • How far from the target the user was?
  • TODO: Was the Decision contested.

Page 2: Analytics Page 3: Master page where we can upload img, add reason + (Target/Range), add link, OR Modify existing img ranges.

Analytics

Trainee Analytics:

  • Nbr of images checked.
  • Fail rate on Good images
  • Fail rate on Bad images
  • Top 3 problematic decision (added or missing).
  • Average inspection time vs other.
  • Graph of success rate split in 100 img since the start.
  • Calender view of his progress (% success vs NbrPer day D3?)

Trainer Analytics:

  • Success rate on Bad images.
  • Success rate on Good images.
  • Images that is mostly mistaken.
  • Images that were Contested by trainee + All reason listed of all user. --> need username + Text + All missing or extra item.
  • Reason that fail the most.
  • Average time for inspection per user.
  • Total system Good img vs Bad img.
  • Total system Nbr Error 1, 2, 3, 4...
  • Calender view of all user progress (% success vs NbrPer day D3?).

Technology

  • Angular2
  • Node backend secured with Passeport.
  • Postgresql
  • D3 for Analytic?

Installation

npm install npm test npm run

Usage

License

MIT

If they fail a img should the img be re-integrated in the system or never showned?