Mission Possible: Migration to WebSphere

sweetleafapartInternet and Web Development

Aug 7, 2012 (4 years and 10 months ago)

278 views

SFSU WebSphere Portal Server Enablement

SFSU WebSphere Portal Server
Enablement

Portal Personalization

Speaker
: Michael Postaski, Senior Portal Developer


email: postaski@us.ibm.com

SFSU WebSphere Portal Server Enablement

What is Personalization?


Customize web content for each user based on profile information


Match users to best content for their interests and needs


Create rules for classifying users and content


Send personalized, targeted emails


Example: show different news articles to managers than to regular
employees, or different information to "gold" customers


Personalization can recognize a specific user based on a profile


Personalization can determine characteristics of a user based on
previous purchases, products or pages viewed, etc.


Personalization then selects content that is appropriate for that profile

SFSU WebSphere Portal Server Enablement

Personalization is composed of


Personalization Browser


Registers resource collections


Authors rules, campaigns, and content spots


Maps rules into content spots for a particular campaign


Rules Engine


Executes rules created in the Personalization browser


LikeMinds Recommendation engine


Evaluates Recommendation rules created in the Personalization browser


Resource engine


Resolves the queries produced by rules into content pieces


Content created using content management tool, SQL, LDAP or any other
database


A Logging framework


Log site usage information

SFSU WebSphere Portal Server Enablement

Type of Personalization


Simple filtering


Display content based on predefined groups



Rules engines


Define business rules to determine what type of content is
displayed for certain types of user profiles



Collaborative filtering


Compare user’s product ratings with ratings of other users
to determine similar content

SFSU WebSphere Portal Server Enablement

How a site is personalized


Develop a personalized portlet or web site


Define an area of a site that needs personalization ex
-

news articles


Define Resource collections and Application Objects


Register these collection and objects to the Personalization server
through the Personalization browser


Use objects and terms defined through the Resource Collections and
Application Objects to write rules and map those rules to content
spots using campaigns


Configure a Personalization Rule Display portlet to show the rule or
content spot


Or call into the Personalization programming interfaces to execute the
appropriate rules or content spots


SFSU WebSphere Portal Server Enablement

How a site is personalized (ctd.)


Execute Personalization rules


User navigates to the page containing Personalization rules or
content spots


Application invokes Personalization to find content or make decisions


Personalization identifies the correct rule(s) to execute


The Personalization server searches for all Rule Mappings for the
Content Spot


Rule Mappings are ordered based on the priority and split values


Rules associated with each mapping are executed in turn until a rule
returns content


For each rule executed, Personalization retrieves the user's profile and
evaluates the rule to select the content which meets the conditions of the
rule


The content is returned to the Web page, and displays for the user

SFSU WebSphere Portal Server Enablement

Personalization terms


resource

-

a Java™ class that defines the properties of a
user or content object


resource instance

-

an instance of the resource class


resource collection

-

a collection of resource instances




resources, resource instances, and resource collections can
be a file systems, an LDAP repositories, an XML store, or
virtually any content store accessible by Java.

SFSU WebSphere Portal Server Enablement

More Personalization terms


rules

-

used to define how your Web site interacts with individual and
groups of Web site visitors. Easy to read logic statements


content spot

-

a placeholder or slot for a rule on a Web page


Each has a unique name


Content type must be defined when it is created and should not be
changed


Created by developers using the Content Spot wizard in Rational
Application Developer and also in the Personalization workspace


rule spot mapping

-

merely an association between a content spot
and a rule


campaigns



a means of organizing and implementing sets of
personalization behavior


Contains sets of rule mappings

SFSU WebSphere Portal Server Enablement

More Personalization terms


Application Object

-

a typed object existing at a known location in the request
context.


specify the object’s type (as a Java class), and a key


Request Context
-

this is the interface used to access various attributes for rules


provides the Personalization rules engine with the data and environmental
information needed for rules processing


contains all the input to execute Personalization content spots and rules


includes simple inputs like request and session data, and more advanced
input like the user object


query framework

-

an object representation of a query


used to pass query information from the runtime engine into the resource engine


It is up to the resource collection to interpret the query

SFSU WebSphere Portal Server Enablement

The Personalization Portlets


Personalization Navigator


Main navigation interface


Users explore the repository with a tree structure


Also control repository content from this portlet


Personalization Editor


Where users edit element content or information


Selecting a new element from the Personalization Navigator activates the
edit mode, where users enter data


Personalized List


Allows a user to display personalized content without having to build a
custom JSP portlet


Each portlet can display a list of resources and show details for each
returned resource


Quickly and easily deploy customized information to users



SFSU WebSphere Portal Server Enablement

The User


The needs and interests of the site visitors are stored as properties in
the user profile data store


Many sites obtain the user needs and preferences by using an HTML
input form


A business might also programmatically update the user profile data
store with information obtained from other sources


Users can be profiled or grouped into categories that facilitate
personalization


Example
-

certain information might be appropriate for managers,
while other content may be suitable for salespersons

SFSU WebSphere Portal Server Enablement

Summary


Content may be dynamically displayed based on user profile
information



Business analysts can determine rules to manage this
process



Content may come from various sources



IBM WebSphere Portal provides a portlet
-
based interface to
manage Personalization artifacts

SFSU WebSphere Portal Server Enablement


QUESTIONS?


SFSU WebSphere Portal Server Enablement