AEM project with content analysis for fragments
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.
core
it.launcher
it.tests
ui.apps
ui.content
README.md
highlight.png
metadata.png
pom.xml
screenshot1.png

README.md

Adobe AEM SmartCrop

Corresponding writeup: SmartCrop

To test the smart crop, retrieve an asset with the specified dimensions, for example:

http://localhost:4502/content/dam/geometrixx/portraits/yolanda_huggins.jpg.100.100.jpg http://localhost:4502/content/dam/geometrixx/portraits/yolanda_huggins.jpg.400.400.jpg

smartcrop example

Adobe AEM Text to Speech

Corresponding writeup: How machine learning can transform content management - ii

content fragment image

Adobe AEM content intelligence with IBM Watson API

Corresponding writeup: How machine learning can transform content management

Outline

Sample project using IBM Watson to analyze content fragments within Adobe AEM. It contains a workflow process steps that calls the Watson API, and assigns the results as metadata to the content fragment. To use, incorporate the workflow process step in a workflow.

The tags are created with namespace bluemix.

An example result using Steve Jobs’ 2005 Stanford commencement speech can be seen in this screenshot: contentintelligence result

Adobe AEM content intelligence with Amazon Comprehend

Corresponding writeup: How machine learning can transform content management - part III

Outline

Sample project using Amazon Comprehend to analyze content fragments within Adobe AEM. It contains a workflow process steps that calls the Comprehend API, and assigns the results as metadata to the content fragment. To use, incorporate the workflow process step in a workflow.

The tags are created with namespace aws.

An example result of metadata assignments using Steve Jobs’ 2005 Stanford commencement speech can be seen in this screenshot: contentintelligence AWS metadata result

And key phrase highlighting: contentintelligence AWS highlight result

Running

For the various integrations, access to AWS or IBM needs to be configured. It will pick up credentials from the environment.

For AWS, this can be done using a Credentials Provider

For IBM, setup access to the IBM Watson Tone Analyzer

Once you retrieve your credentials, set the following environment variables.

    export IBM_API_USER_NAME=<api_user_name>
    export IBM_API_PASSWORD=<api_password>
  • Start your AEM instance in this environment.
  • Create a workflow to use the tone analyzer process steps and select an asset.

Project Structure

This is created based on the standard project template for AEM-based applications.

How to build

To build all the modules run in the project root directory the following command with Maven 3:

mvn clean install

If you have a running AEM instance you can build and package the whole project and deploy into AEM with

mvn clean install -PautoInstallPackage

Or to deploy it to a publish instance, run

mvn clean install -PautoInstallPackagePublish

Or to deploy only the bundle to the author, run

mvn clean install -PautoInstallBundle

Testing

There are three levels of testing contained in the project:

  • unit test in core: this show-cases classic unit testing of the code contained in the bundle. To test, execute:

    mvn clean test

  • server-side integration tests: this allows to run unit-like tests in the AEM-environment, ie on the AEM server. To test, execute:

    mvn clean integration-test -PintegrationTests

  • client-side Hobbes.js tests: JavaScript-based browser-side tests that verify browser-side behavior. To test:

    in the browser, open the page in 'Developer mode', open the left panel and switch to the 'Tests' tab and find the generated 'MyName Tests' and run them.

Maven settings

The project comes with the auto-public repository configured. To setup the repository in your Maven settings, refer to:

http://helpx.adobe.com/experience-manager/kb/SetUpTheAdobeMavenRepository.html