Skip Headers
Oracle® Real-Time Decisions Base Application Installation and Reference Guide
Release 3.1

Part Number E19020-01
Go to Documentation Home
Home
Go to Book List
Book List
Go to Table of Contents
Contents
Go to Feedback page
Contact Us

Go to previous page
Previous
Go to next page
Next
View PDF

4 Base E-Commerce

This chapter describes the elements in the Inline Service RTD_Base_ECommerce. It contains the following topics:

4.1 Introduction to Base E-Commerce

The following diagram shows an overview of the stages of a typical ECommerce application and its possible interactions with the Inline Service RTD_Base_ECommerce.

Surrounding text describes ecomm_integration_points.gif.

The RTD_Base_ECommerce Inline Service serves as a general ECommerce framework for customers to adapt to their business processes.

The RTD_Base_ECommerce Inline Service provides pre-defined entities, choices, decisions, models, and integration points.

From an end user perspective, the RTD_Base_ECommerce Inline Service is designed on the assumption that customers will perform additional configuration and further customization to fulfill their business requirements.

Additional tasks involve mapping the logical entity attributes to a customer's physical data sources and to develop the front-end environment for the presentation of any Oracle RTD recommendations.

For more information, see Chapter 6, "Configuring the Base Inline Services."

4.2 Overview of Choices Groups and Choices

This section presents an overview of the choice groups and choices, and their usage in the integration points.

This section contains the following topics:

For more information about how the choices and choice groups are used, see Section 4.5, "Advisors and Associated Decisions, Choice Groups, and Models."

4.2.1 Offer-Oriented Choices

Choices for offer-oriented use cases are structured as follows:

Surrounding text describes choices_decision.gif.

Offers and all other choice groups under it are comprised of dynamic choices. The choice data is either supplied as an advisor's incoming parameter value (see Section 4.5, "Advisors and Associated Decisions, Choice Groups, and Models") or retrieved by Oracle RTD from external data sources.

Advisors for Offer-Oriented Choices

The offer-oriented choices are returned when invoking the following advisors:

  • Get Cross Sell Offers

  • Get Upsell Offers

  • Get Promotions

  • Get Advertisements

  • Get Offers

    • Updates current web interaction attributes and Web Pages for the session.

    • Identifies any given page a customer navigates to that needs to be tracked.

For more specific information about the advisors, see Section 4.5, "Advisors and Associated Decisions, Choice Groups, and Models."

4.2.2 Interaction-Oriented Choices

Choice groups and choices for interaction-oriented usages are structured as follows. Note that the choices under each choice group can be reconfigured to suit the end users needs and serve as a template.

Surrounding text describes choices_analysis.gif.
  • Abandonment

    • Identifies if customers have abandoned their web session. This would be determined by the business based on their definitions of abandonment.

      Abandonment choice can be obtained by invoking the advisor Get Abandonment Propensity, and its prediction model is updated by invoking the advisor Close Session.

    • Returns likelihood of abandonment for a customer.

  • Web Actions

    • Identifies key actions done by a customer that can later be applied to entity attributes or models, for example, Update Customer Profile, Register, Un-register. Chat request.

      Web Actions choice can be obtained by invoking the advisor Get Likely Web Action, and its prediction model is updated by invoking the informant Customer Action.

  • Web Site Duration

    • Web Site Durations's prediction model is updated by invoking the informant Close Session.

  • Web Support Types

    • Web Support Types choice can be obtained by invoking the advisor Get Web Support Types, and its prediction model is updated by invoking the informant Web Support Feedback.

    • Returns likelihood for a customer to accept chat request or email support if offered.

4.2.3 Informants Usage with Offer-Oriented and Transaction-Oriented Choices

This section describes the following usages of the RTD_Base_ECommerce Inline Service informants:

For more specific information about the informants, see Section 4.4, "Informants."

  • Initiate Session

    • Updates current web interaction attributes for the session.

  • Identify Customer

    • Updates customer related attributes for the session.

    • Provides Oracle RTD with customer id am profile information if a customer is identified while on the web site. In most cases, identification can be made upon user logon or by previously set cookies at the customer's browser.

  • Page Turn

    • Updates current web interaction attributes and Web Page for the session.

    • Identifies any given page a customer navigates to that needs to be tracked.

  • Performed Search

    • Updates current web interaction attributes and Search keywords for the session.

    • Identifies search words used by a customer.

  • Customer Action

    • Updates current web interaction attributes and Web Action for the session.

    • Identifies key actions performed by a customer that can later be applied to entity attributes or models, for example, Update Customer Profile, Register, Un-register, Chat request.

  • Added To Cart

    • Updates current web interaction attributes and Shopping Cart Items for the session.

  • Offer Response

    • Creates prediction model entries and also captures shopping cart addition events.

    • Identifies if an offer (Upsell, Cross Sell, Promotion, or Ad) is clicked, added to a cart, or purchased by a web user or customer.

  • Close Session

    • Creates analytical prediction model entries and triggers learning.

    • Formally closes out the Oracle RTD session.

4.3 Entities

The following diagram illustrates the overall general logical entity object model of the RTD_Base_ECommerce Inline Service.

Surrounding text describes base_ecom_logical_model.gif.

The following diagram shows the entities and the relationships between the entities defined in the RTD_Base_ECommerce Inline Service.

Surrounding text describes rtd_entity_model.gif.

Notation

The diagram shows the standard notation used in UML class diagrams, with each directed line representing a relationship from element A to element B, as follows:

The multiplicity of a relationship restricts how many element B instances the relationship may have. The restriction denotes either a precise limit, such as 1 or 0..1, or an open-ended upper limit, such as "zero or more" or "one or more."

For example:

System-Oriented Entities

The Session entity is a built-in Oracle RTD entity for maintaining session attribute values.

The following entities are used for dynamic choice retrieval from external data sources: Ad List, Promotion List, Cross Sell Product List, and Up Sell Product List.

Entities Outline

The Current Web Interaction entity is a session attribute that keeps track of the current user interaction with the client system.

The Current Web Interaction entity references the Web Interaction entity, that itself keeps track of the following data:

The Customer entity is a session attribute that contains details of the customer profile as well as past customer interaction behavior.

A customer can be a Person or an Organization.

The Customer may have historical information, in the entities Purchase History, Campaign History, and Interaction History. Past customer interactions could be either or both of the following:

After a Customer Id has been identified and supplied to Oracle RTD by the Identify Customer informant, Oracle RTD retrieves data from external data sources for the Customer entity and for its associated entity attributes.

In addition to the entity attributes that are normally mapped to a data source directly, derived attributes are also included, which obtain their values via Java functions that utilize the applicable raw data extracted from data sources as inputs.

This section describes the following entities:

4.3.1 Session Entity

(Key = Web Session Id)

Table 4-1 Session Entity

Attribute Array Type Show In DC Use for Analysis Comments

Customer

No

Customer

Yes

Yes

None

Current Web Interaction

No

Current Web Interaction

Yes

Yes

None

Rank Offers

No

Rank Offers

No

No

Utilized when ranking offers directly from an Offer advisor instead of read from a data source.

Supplied Product Id

No

String

No

No

Utilized when ranking offers directly from an Offer advisor instead of read from a data source.


4.3.2 Ad

The Ad entity is used in conjunction with the Ad List entity and is used for the dynamic choice associated with the Ad Choice group.

Table 4-2 Ad Entity

Attribute Array Type Show In DC Use for Analysis Comments

Ad Id

No

String

No

No

Key

Category

No

String

No

No

None

Name

No

String

No

No

None

Type

No

String

No

No

None


4.3.3 Ad List

The Ad List entity is used in conjunction with the Ad entity and is used for the dynamic choice associated with the Ad Choice group.

Table 4-3 Ad List Entity

Attribute Array Type Show In DC Use for Analysis Comments

Active

No

String

No

No

Key

Ads

Yes

Ad

No

No

The Ads attribute is an array attribute based on the Ad entity.


4.3.4 Agent Interaction

The Agent Interaction entity contains attributes related to agent interactions that have taken place with a customer. This entity is used to create an array of agent interactions within the Interaction History entity, which in turn is associated with the Customer entity.

Table 4-4 Agent Interaction Entity

Attribute Array Type Show In DC Use for Analysis Comments

Interaction Id

No

String

No

No

None

Agent Id

No

String

No

No

None

Agent Location

No

String

No

No

None

Customer Location

No

String

No

No

None

Interaction Channel

No

String

No

No

None

Interaction Date

No

None

No

No

None

Interaction Duration

No

Integer

No

No

None

Interaction Reason

No

String

No

No

None

Interaction Status

No

String

No

No

None

Interaction Type

No

String

No

No

None

Time Of Day

No

String

No

No

None


4.3.5 Campaign

The Campaign entity contains attributes related to campaigns that have been associated with a customer. This entity is associated with the customer via the Campaign Item and Campaign History entities.

Table 4-5 Campaign Entity

Attribute Array Type Show In DC Use for Analysis Comments

Campaign Id

No

String

No

No

None

Category

No

String

Yes

Yes

None

Name

No

String

Yes

Yes

None

Period

No

String

Yes

Yes

None

Type

No

String

Yes

Yes

None


4.3.6 Campaign History

The Campaign History entity contains attributes related to campaigns that have been associated with a customer. This entity is associated with the Customer entity and also contains the Campaign Items entity.

Table 4-6 Campaign History Entity

Attribute Array Type Show In DC Use for Analysis Comments

Last Campaign Category

No

String

Yes

Yes

Default Value based on the Get Last Campaign function

Last Campaign Date

No

Date

Yes

Yes

Default Value based on the Get Last Campaign Date function

Last Campaign Delivery Method

No

String

Yes

Yes

Default Value based on the Get Last Campaign Delivery Method function

Last Campaign Name

No

String

Yes

Yes

Default Value based on the Get Last Campaign Name function

Last Campaign Type

No

String

Yes

Yes

Default Value based on the Get Last Campaign Type function

Days Since Last Campaign

No

Integer

Yes

Yes

Default Value based on the Get Days Since Last Campaign

Campaign Items

Yes

Campaign Item

Yes

Yes

The Campaign Items attribute is an array attribute based on the Campaign Item entity.


4.3.7 Campaign Item

The Campaign Item entity contains attributes related to campaigns that have been associated with a customer. This entity is associated with the customer via the Campaign History entity and contains the Campaign entity.

Table 4-7 Campaign Item Entity

Attribute Array Type Show In DC Use for Analysis Comments

Campaign Date

No

Date

Yes

Yes

None

Delivery Method

No

String

Yes

Yes

None

Campaign

No

Campaign

Yes

Yes

The Campaign attribute is based on the Campaign entity.


4.3.8 Cart Item

The Cart Item entity contains attributes related to the products that a customer has put in their shopping cart in their web session. The Cart Item entity contains the Product entity, and is itself embedded in the Web Interaction entity.

Table 4-8 Cart Item Entity

Attribute Array Type Show In DC Use for Analysis Comments

Quantity

No

Integer

Yes

Yes

None

Product

No

Product

Yes

Yes

None


4.3.9 Clicked Ad

The Clicked Ad entity contains attributes related to the ads which a customer may have clicked during their web session. The Clicked Ad entity contains the Ad entity, and is itself embedded in the Web Interaction entity.

Table 4-9 Clicked Ad Entity

Attribute Array Type Show In DC Use for Analysis Comments

Count

No

Integer

Yes

Yes

None

Ad

No

Ad

Yes

Yes

None


4.3.10 Clicked Promotion

The Clicked Promotion entity contains attributes related to the ads which a customer may have clicked during their web session. The Clicked Promotion entity contains the Promotion entity, and is itself embedded in the Web Interaction entity.

Table 4-10 Clicked Promotion Entity

Attribute Array Type Show In DC Use for Analysis Comments

Count

No

Integer

Yes

Yes

None

Promotion

No

Promotion

Yes

Yes

None


4.3.11 Cross Sell Product List

The Cross Sell Product List entity is used in conjunction with the Products entity (instantiated as Cross Sell Products), and is used for the dynamic choice associated with the Cross Sell Offers Choice group.

Table 4-11 Cross Sell Product List Entity

Attribute Array Type Show In DC Use for Analysis Comments

Product Id

No

String

No

No

None

Cross Sell Products

Yes

Product

No

No

The Cross Sell Products attribute is an array attribute based on the Product entity.

Cross Sell Products - Product Id

No

String

No

No

None


4.3.12 Current Web Interaction

The Current Web Interaction entity contains attributes related to what a customer is doing during their current web session. The Current Web Interaction entity contains its own attributes as well as attributes from the Web Interaction entity.

See Section 4.3.27.1, "Derivation of Web Interaction Attributes in Referencing Entities" for details of how the Web Interaction attributes Interaction Date, Start Time, and Total Duration in Minutes are derived.

Table 4-12 Current Web Interaction Entity

Attribute Array Type Show In DC Use for Analysis Comments

Interaction Id

No

String

Yes

Yes

None

Current Page

No

String

Yes

Yes

None

Current Page Type

No

String

Yes

Yes

None

Current Viewed Product Id

No

String

Yes

Yes

None

Last Added Product Id

No

String

Yes

Yes

None

Previous Page

No

String

Yes

Yes

None

Previous Page Type

No

String

Yes

Yes

None

Time Spent on Previous Page

No

Integer

Yes

Yes

None

Web Interaction

No

Web Interaction

Yes

Yes

None


4.3.13 Customer

The Customer entity contains attributes related to the profile of the customer. The Customer entity contains its own attributes and links in the Campaign History, Interaction History, Organization, Person, and Purchase History entities.

Table 4-13 Customer Entity

Attribute Array Type Show In DC Use for Analysis Comments

Customer Id

No

String

No

No

Key

Address City

No

String

Yes

Yes

None

Address Country

No

String

Yes

Yes

None

Address Postal Code

No

String

Yes

Yes

None

Address Region

No

String

Yes

Yes

None

Address State Province

No

String

Yes

Yes

None

Credit Hold

No

String

Yes

Yes

None

Life Time Value Score

No

Double

Yes

Yes

None

Offline Churn Propensity

No

Double

Yes

Yes

None

Phone Area Code

No

String

Yes

Yes

None

Preferred Language

No

String

Yes

Yes

None

Start Date

No

Date

Yes

Yes

None

Status

No

String

Yes

Yes

None

Target Market Segment

No

String

Yes

Yes

None

Tenure

No

Integer

Yes

Yes

None

Total Credit Limit

No

Double

Yes

Yes

None

Type

No

String

Yes

Yes

Default Value determined by Get Customer Type function.

Campaign History

No

Campaign History

Yes

Yes

The Campaign History attribute is based on the Campaign History entity. See Campaign History entity for attribute breakdown.

Interaction History

No

Interaction History

Yes

Yes

The Interaction History attribute is based on the Interaction History entity. See Interaction History entity for attribute breakdown.

Organization

No

Organization

Yes

Yes

The Organization attribute is based on the Organization entity. See Organization entity for attribute breakdown.

Person

No

Person

Yes

Yes

The Person attribute is based on the Person entity. See Person entity for attribute breakdown.

Purchase History

No

Purchase History

Yes

Yes

The Purchase History attribute is based on the Purchase History entity. See Purchase History entity for attribute breakdown.


4.3.14 Interaction History

The Interaction History entity contains attributes that record what a customer has done in the past regarding previous interactions. The Interaction History entity contains derived attributes from both previous Agent Interactions and previous Web Interactions.

See Section 4.3.27.1, "Derivation of Web Interaction Attributes in Referencing Entities" for details of how the Past Web Interactions attributes Interaction Date, Start Time, and Total Duration in Minutes are derived.

Table 4-14 Interaction History Entity

Attribute Array Type Show In DC Use for Analysis Comments

Agent Interaction Reasons In Past 30 Days

Yes

String

Yes

Yes

Default determined by Get Agent Interaction Reasons In Past Days function

Agent Interaction Types in Past 30 Days

Yes

String

Yes

Yes

Default determined by Get Agent Interaction Types In Past Days function

Days Since Last Agent Interaction

No

Integer

Yes

Yes

Default determined by Get Days Since Last Agent Interaction function

Days Since Last Interaction

No

Integer

Yes

Yes

Default determined by Maximum function

Days Since Last Web Interaction

No

Integer

Yes

Yes

Default determined by Get Days Since Last Web Interaction function

Interaction Types In Past 30 Days

No

String

Yes

Yes

Default determined by Get Interaction Types function

Last Agent Interaction Status

No

String

Yes

Yes

Default determined by Get Last Agent Interaction Status function

Last Agent Interaction Type

No

String

Yes

Yes

Default determined by Get Last Agent Interaction Type function

Last Interaction Type

No

String

Yes

Yes

Default determined by Get Last Interaction Type function

Number of Agent Interactions In Past 30 Days

No

Integer

Yes

Yes

Default determined by Get Number Of Agent Interaction in Past Days function

Number of Agent Interactions In Past 90 Days

No

Integer

Yes

Yes

Default determined by Get Number Of Agent Interaction in Past Days function

Number of Agent Interactions In Past Year

No

Integer

Yes

Yes

Default determined by Get Number Of Agent Interaction in Past Days function

Number of Web Interactions In Past 30 Days

No

Integer

Yes

Yes

Default determined by Get Number Of Web Interaction in Past Days function

Number of Web Interactions In Past 90 Days

No

Integer

Yes

Yes

Default determined by Get Number Of Web Interaction in Past Days function

Number of Web Interactions In Past Year

No

Integer

Yes

Yes

Default determined by Get Number Of Web Interaction in Past Days function

Performed Web Actions In Past 30 Days

Yes

String

Yes

Yes

Default determined by Get Performed Web Actions in Past Days function

Past Agent Interactions

Yes

Agent Interaction

No

No

None

Past Web Interactions

Yes

Web Interaction

No

No

None


4.3.15 Organization

The Organization entity contains attributes related to the profile of an Organization. The Organization entity is linked to the session through the Customer entity.

Table 4-15 Organization Entity

Attribute Array Type Show In DC Use for Analysis Comments

Annual Gross Profit

No

Double

Yes

Yes

None

Annual Revenue

No

Double

Yes

Yes

None

Business Partner Flag

No

String

Yes

Yes

None

Established Service

No

Integer

Yes

Yes

None

Line Of Business

No

String

Yes

Yes

None

Number of Employees

No

Integer

Yes

Yes

None

Number Of Years Established

No

Integer

Yes

Yes

None

Size

No

String

Yes

Yes

None

Type

No

String

Yes

Yes

None


4.3.16 Person

The Person entity contains attributes related to the profile of a Person. The Person entity is linked to the session through the Customer entity.

Table 4-16 Person Entity

Attribute Array Type Show In DC Use for Analysis Comments

Age

No

Integer

Yes

Yes

None

Annual Income

No

Double

Yes

Yes

None

Credit Score

No

Integer

Yes

Yes

None

Education Level

No

String

Yes

Yes

None

Ethnicity

No

String

Yes

Yes

None

Gender

No

String

Yes

Yes

None

Marital Status

No

String

Yes

Yes

None

Net Worth

No

Double

Yes

Yes

None

Number Of Children

No

Integer

Yes

Yes

None

Profession

No

String

Yes

Yes

None


4.3.17 Product

The Product entity contains attributes related to a generic Product. The Product entity is used as a reference entity under the Cart Item, Cross Sell Product List, Promoted Item, Purchased Item, and Up Sell Product entities.

Table 4-17 Product Entity

Attribute Array Type Show In DC Use for Analysis Comments

Product Id

No

String

No

No

Key

Category

No

String

Yes

Yes

None

Name

No

String

Yes

Yes

None

Popularity Rank

No

Integer

Yes

Yes

None

Product Line

No

String

Yes

Yes

None

Type

No

String

Yes

Yes

None

Unit Price

No

Double

Yes

Yes

None


4.3.18 Promoted Item

The Promoted Item entity contains attributes related to a Promoted Item. The Promoted Item entity is used in conjunction with the Promotion entity, which can contain multiple promoted Items.

Table 4-18 Promoted Item Entity

Attribute Array Type Show In DC Use for Analysis Comments

Discount Rate

No

Double

Yes

Yes

None

Promoted Product

No

Product

Yes

Yes

None


4.3.19 Promotion

The Promotion entity is used in conjunction with the Promotion List entity, and is used for the dynamic choice associated with the Promotions Choice group.

Table 4-19 Promotion Entity

Attribute Array Type Show In DC Use for Analysis Comments

Promotion Id

No

String

No

No

None

Category

No

String

No

No

None

Days Left

No

Integer

No

No

Default Value set by Get Days Left function

Duration In Days

No

Integer

No

No

Default Value set by Get Duration In Days function

Effective Date

No

Date

No

No

None

Expiry Date

No

Date

No

No

None

Name

No

String

No

No

None

Period

No

String

No

No

None

Type

No

String

No

No

None

Promoted Items

Yes

Promoted Item

No

No

The Promoted Items attribute is based on the Promoted Item entity. See Promoted Item entity for attribute breakdown.


4.3.20 Promotion List

The Promotion List entity is used in conjunction with the Promotion entity, and is used for the dynamic choice associated with the Promotions Choice group.

Table 4-20 Promotion List Entity

Attribute Array Type Show In DC Use for Analysis Comments

Active

No

String

No

No

Key

Promotions

Yes

Promotion

No

No

The Promotions attribute is based on the Promotion entity. See Promotion entity for attribute breakdown.


4.3.21 Purchased Item

The Purchased Item entity contains attributes related to a purchased item. This entity is associated with the Customer entity through the Purchase History entity.

Table 4-21 Purchased Item Entity

Attribute Array Type Show In DC Use for Analysis Comments

Purchase Amount

No

Double

Yes

Yes

None

Purchase Date

No

Date

Yes

Yes

None

Purchased Product

No

Product

Yes

Yes

None


4.3.22 Purchase History

The Purchased History entity contains attributes related to a customer's past purchases. This entity is associated with the session via the Customer entity.

Table 4-22 Purchase History Entity

Attribute Array Type Show In DC Use for Analysis Comments

Days Since Last Purchase

No

Integer

Yes

Yes

Default Value set by Get Days Since Last Purchase function

Last Purchase Amount

No

Double

Yes

Yes

Default Value set by Get Last Purchase Amount function

Last Purchased Product

No

String

Yes

Yes

Default Value set by Get Last Purchased Product function

Last Purchased Product Line

No

String

Yes

Yes

Default Value set by Get Last Purchased Product Line function

Product Lines Owned

Yes

String

Yes

Yes

Default Value set by Get Product Lines Owned function

Total Amount Spent

No

Double

Yes

Yes

Default Value set by Get Total Amount Spent function

Total Amount Spent in Last 90 Days

No

Double

Yes

Yes

Default Value set by Get Total Amount Spent In Last 90 Days function

Purchased Items

Yes

Purchased Item

Yes

Yes

The Purchased Items array attribute is based on the Purchased Item entity. See Purchased Item entity for attribute breakdown.


4.3.23 Rank Offers

The Rank Offers entity is used to store arrays of different offer types that are passed to the Inline Service via an advisor. After it is filled by the advisor input, the entity is then used to populate the corresponding dynamic choice.

Table 4-23 Rank Offers Entity

Attribute Array Type Show In DC Use for Analysis Comments

Ads

Yes

String

No

No

None

CrossSellOffers

Yes

String

No

No

None

In Use

No

Boolean

No

No

Default Value is False

Promotions

Yes

String

No

No

 

UpSellOffers

Yes

String

No

No

 

4.3.24 Search

The Search entity is used to store attributes related to search strings performed by a customer. This entity is used by the Web Interaction entity as an array attribute.

Table 4-24 Search Entity

Attribute Array Type Show In DC Use for Analysis Comments

Count

No

Integer

Yes

Yes

None

Keyword

No

String

Yes

Yes

None


4.3.25 Up Sell Product List

The Up Sell Product List entity is used in conjunction with the Products entity (instantiated as Up Sell Products), and is used for the dynamic choice associated with the Up Sell Offers Choice group.

Table 4-25 Up Sell Product List Entity

Attribute Array Type Show In DC Use for Analysis Comments

Product Id

No

String

No

No

None

Up Sell Products

Yes

Product

No

No

The Up Sell Products attribute is an array attribute based on the Product entity.


4.3.26 Web Action

The Web Action entity is used to store attributes related to actions performed by a customer while on the web. This entity is used by the Web Interaction entity as an array attribute.

Table 4-26 Web Action Entity

Attribute Array Type Show In DC Use for Analysis Comments

Name

No

String

No

No

Key

Category

No

String

Yes

Yes

None

Type

No

String

Yes

Yes

None

Web Action Id

No

String

Yes

Yes

None


4.3.27 Web Interaction

The Web Interaction entity contains attributes related to a generic Web Interaction. This entity is used as a reference entity under for the Current Web Interaction entity as well as an array attribute (Past Web Interactions) under the Interaction History entity.

See Section 4.3.27.1, "Derivation of Web Interaction Attributes in Referencing Entities" for details of how the Interaction Date, Start Time, and Total Duration in Minutes attributes are used in the Current Web Interaction entity and the Interaction History entity.

Table 4-27 Web Interaction Entity

Attribute Array Type Show In DC Use for Analysis Comments

Interaction Id

No

String

No

No

Key

Clicked Ads

Yes

Clicked Ad

Yes

Yes

The Clicked Ads attribute is an array attribute based on the Clicked Ad entity. See the Clicked Ad entity for attribute breakdown.

Clicked Promotions

Yes

Clicked Promotion

Yes

Yes

The Clicked Promotions attribute is an array attribute based on the Clicked Promotion entity. See the Clicked Promotion entity for attribute breakdown.

Interaction Date

No

Date

Yes

Yes

None

Origin To Website

No

String

Yes

Yes

None

Performed Actions

Yes

Web Action

Yes

Yes

The Performed Actions attribute is an array attribute based on the Web Action entity. See the Web Action entity for attribute breakdown.

Performed Searches

Yes

Search

Yes

Yes

The Performed Searches attribute is an array attribute based on the Search entity. See the Search entity for attribute breakdown.

Shopping Cart Items

Yes

Cart Item

Yes

Yes

The Shopping Cart Items attribute is an array attribute based on the Cart Item entity. See the Cart Item entity for attribute breakdown.

Start Time

No

Date

No

No

Attribute used to derive the Web Interaction attribute Total Duration in Minutes when used in the Current Web Interaction entity.

Time Of Day

No

String

Yes

Yes

None

Total Duration in Minutes

No

String

Yes

Yes

None

Visited Pages

Yes

Web Page

Yes

Yes

The Visited Pages attribute is an array attribute based on the Web Page entity. See the Web Page entity for attribute breakdown.

Web User Location

No

String

Yes

Yes

None


4.3.27.1 Derivation of Web Interaction Attributes in Referencing Entities

When the Current Web Interaction and the Interaction History reference the Web Interaction entity, special considerations apply to the derivation of certain attributes, as follows:

Table 4-28 Derivation of Web Interaction Attributes in Referencing Entities

Web Interaction Attribute As used in the Current Web Interaction Entity attribute Web Interaction As used in the Interaction History Entity attribute Past Web Interactions

Interaction Date

NA

Retrieved from the external data source using Web Interaction mapping

Start Date

Initialized by informant Initiate Session with the Current Time

Not initialized

Total Duration in Minutes

Computed as the difference between Web Interaction-Start Time and Current Time

Retrieved from the external data source using Web Interaction mapping


4.3.28 Web Page

The Web Page entity is used to store attributes related to the attributes of a web page that a customer has visited while on the web. This entity is used by the Web Interaction entity as an array attribute.

Table 4-29 Web Page Entity

Attribute Array Type Show In DC Use for Analysis Comments

Web Page Id

No

String

No

No

Key

Average Time Spent On Page

No

Double

Yes

Yes

None

Category

No

String

Yes

Yes

None

Count

No

Integer

Yes

Yes

None

Name

No

String

Yes

Yes

None

Type

No

String

Yes

Yes

None


4.4 Informants

This section describes the following Informants:

4.4.1 Initiate Session

The informant Initiate Session creates the session for the interaction and updates current web interaction attributes for the session.

Table 4-30 describes the parameters for the informant Initiate Session.

Table 4-30 Informant Initiate Session

Parameter Description

Informant Name

Initiate Session

Session Keys

Session / Current Web Interaction / Interaction Id

Request Data

Origin To Website => Session / Current Web Interaction / Web Interaction / Origin To Website

Time Of Day => Session / Current Web Interaction / Web Interaction / Time Of Day

User Location => Session / Current Web Interaction / Web Interaction / Web User Location

External System

Web E-Commerce

Order

0

Force session close

No

Logic

None

Pre-condition

None


4.4.2 Identify Customer

The informant Identify Customer updates and triggers loading of the following customer related attributes for the session:

  • Customer Profile

  • Interaction History

  • Purchase History

  • Campaign History

The caller provides Oracle RTD with the Customer Id after a customer is identified while on the web site. In most cases, identification can be made upon user logon or by previously set cookies on the caller's web browser.

Table 4-31 describes the parameters for the informant Identify Customer.

Table 4-31 Informant Identify Customer

Parameter Description

Informant Name

Identify Customer

Session Keys

Session / Current Web Interaction / Interaction Id

Request Data

Customer Id => Customer / Customer Id

External System

Web E-Commerce

Order

0

Force session close

No

Logic

None

Pre-condition

None


4.4.3 Page Turn

The informant Page Turn updates current web interaction attributes (Web Pages) for the session. Identifies any given page a customer navigates to that needs to be tracked

Table 4-32 describes the parameters for the informant Page Turn.

Table 4-32 Informant Page Turn

Parameter Description

Informant Name

Page Turn

Session Keys

Session / Current Web Interaction / Interaction Id

Request Data

Current Page => Session / Current Web Interaction / Current Page

Current Page Type => Session / Current Web Interaction / Current Page Type

Previous Page => Session / Current Web Interaction / Previous Page

Previous Page Type => Session / Current Web Interaction / Previous Page Type

Time Spent on Previous Page => Session / Current Web Interaction / Time Spent on Previous Page

Current Viewed Product Id => Session / Current Web Interaction / Current Viewed Product Id

External System

Web E-Commerce

Order

0

Force session close

No

Logic

(Asynchronous) Update the nested Web Pages session attributes

Pre-condition

None


Supplied Current Viewed Product Id can be used later as the base product id in making cross sell or up sell decision.

4.4.4 Performed Search

The informant Performed Search updates current web interaction attributes (Search Keywords) for the session. It identifies search words used by a user or customer.

Table 4-33 describes the parameters for the informant Performed Search.

Table 4-33 Informant Performed Search

Parameter Description

Informant Name

Performed Search

Session Keys

Session / Current Web Interaction / Interaction Id

Request Data

Search Keyword

External System

Web E-Commerce

Order

0

Force session close

No

Logic

(Asynchronous) Update the nested Searches session attributes

Pre-condition

None


4.4.5 Customer Action

The informant Customer Action updates current web interaction attributes (Web Action) for the session. It identifies key actions performed by a customer that can later be applied to models, for example, Update Customer Profile, Register, Un-register, Chat request.

Table 4-34 describes the parameters for the informant Customer Action.

Table 4-34 Informant Customer Action

Parameter Description

Informant Name

Customer Action

Session Keys

Session / Current Web Interaction / Interaction Id

Request Data

Action Name

External System

Web E-Commerce

Order

0

Force session close

No

Logic

Update Web Action Model (choice model) Create or Update Web Action session variable of current interaction

Pre-condition

None


Set Action Name as a choice name into the Web Action Model.

4.4.6 Added To Cart

The informant Added To Cart updates current web interaction attributes (Cart Item) for the session. It registers a shopping cart item along with its quantity added during the session.

Table 4-35 describes the parameters for the informant Added To Cart.

Table 4-35 Informant Added to Cart

Parameter Description

Informant Name

Customer Action

Session Keys

Session / Current Web Interaction / Interaction Id

Request Data

Product Id Quantity

External System

Web E-Commerce

Order

0

Force session close

No

Logic

Update Web Action Model (choice model) Create or Update Shopping Cart Item session variable of current interaction

Pre-condition

None


4.4.7 Offer Response

The informant Offer Response creates prediction model entries and also captures shopping cart addition events. It identifies if an offer (Up Sell, Cross Sell, Promotion, Ad) is clicked, added to cart, or purchased by a web user or customer.

Table 4-36 describes the parameters for the informant Offer Response.

Table 4-36 Informant Offer Response

Parameter Description

Informant Name

Offer Response

Session Keys

Session / Current Web Interaction / Interaction Id

Request Data

Offer Id {choice id}

Offer Type {Up Sell, Cross Sell, Promotion, Ad}

Event {Interested, Added To Cart, Purchased}

External System

Web E-Commerce

Order

0

Force session close

No

Logic

If event = Interested then if offer type = Promotion then create/update Clicked Promotion (increase count if previously exist)

if offer type = Ad then create/update Clicked Ad (increase count if previously exist)

Record the event into the appropriate model for the choice (determined by choice id / offer id)

Pre-condition

None


4.4.8 Web Support Feedback

The informant Web Support Feedback updates the Web Support prediction model.

Table 4-37 describes the parameters for the informant Web Support Feedback.

Table 4-37 Informant Web Support Feedback

Parameter Description

Informant Name

Web Support Feedback

Session Keys

Session / Current Web Interaction / Interaction Id

Request Data

Choice Name

Choice Event

External System

Web E-Commerce

Order

0

Force session close

No

Logic

If Choice Event = Used then record the event into the Web Support Usage Model

Pre-condition

None


Dependencies:

  • Function - Set Choice Event Model (Model Name, Choice Name, Choice Event)

4.4.9 Close Session

The informant Close Session creates analytical prediction model entries, triggers learning, and formally closes out the Oracle RTD session.

Table 4-38 describes the parameters for the informant Close Session.

Table 4-38 Informant Close Session

Parameter Description

Informant Name

Close Session

Session Keys

Session / Current Web Interaction / Interaction Id

Request Data

Abandonment Flag

Interaction Duration

External System

Web E-Commerce

Order

0

Force session close

No

Logic

Update Abandonment Model.

Possible values for Abandonment Flag are:

"Abandoned"

"Not Abandoned"

Update Web Site Duration Model.

Possible values for Interaction Duration are:

"00 to 05 Minutes"

"05 to 10 Minutes"

"10 to 15 Minutes"

"15 to 20 Minutes"

"20 Minutes or Greater"

Pre-condition

None


Dependencies:

  • Function - Set Choice Model (Model Name, Choice Name)

4.5 Advisors and Associated Decisions, Choice Groups, and Models

For each advisor listed in this section, a detailed breakdown is provided for the Integration Point, followed by:

This section consists of the following topics:

4.5.1 Decisioning Advisors

This section describes the following advisors:

4.5.1.1 Get Upsell Offers

The advisor Get Upsell Offers determines the likelihood for a customer to accept an upsell offer.

Table 4-39 describes the parameters for the advisor Get Upsell Offers.

Table 4-39 Advisor Get Upsell Offers

Parameter Description

Advisor Name

Get Upsell Offers

Session Keys

Session / Current Web Interaction / Interaction Id

Request Data

Product Id (optional)

Number of Offers

Rank Offers (array, optional, to be ranked, as source for dynamic choice)

External System

Web E-Commerce

Order

0

Force session close

No

Decision

Select Upsell Offers

Group Decision

Select Upsell Offers

Default Choices

None

Logic

The supplied product id is assigned to Session / Supplied Product Id. This will be used by function: Select Product Id (for Cross Sell / Up Sell).

If supplied, override the default number of choices returned by decision using the Number of Offers incoming parameter.

If Rank Offer (string array) is supplied, set these into session as dynamic choice objects to be used by decision later.

Pre-condition

Base Product Id for Cross sell / Up sell must exist. Function: Select Product Id tries to find this Product Id from the following variables:

1 - Session / Supplied Product Id

2 - Session / Current Web Interaction / Current Viewed Product Id

3 - Session / Current Web Interaction / Last Added (into shopping cart) Product Id


Table 4-40 describes the parameters for the decision for the advisor Get Upsell Offers.

Table 4-40 Decision for Advisor Get Upsell Offers

Parameter Description

Decision Name

Select Upsell Offers

Select Choices From

Upsell Offers

Number of Choices to Select

5

Select at Random

No

Target Segments

Default

Priorities for Default Segment

Maximize Acceptance Likelihood 33%

Maximize Revenue 33%

Popularity 33%

Pre Selection Logic

None

Post Selection Logic

Generate learning entries

for (int i = 0; i < choices.size(); i++) {choices.get(i).recordEvent("Presented");}


Table 4-41 describes the parameters for the choice group for the advisor Get Upsell Offers.

Table 4-41 Choice Group for Advisor Get Upsell Offers

Parameter Description

Choice Group Name

Upsell Offers

Choice Attributes

Product Id

Product Name

Product Line

Product Type

Product Category

Offer Type = "Up Sell"

Likelihood

Scores

Maximize Acceptance Likelihood = Predicted by Upsell Purchase Model: Purchased

Maximize Revenue = Dynamic Choice / Unit Price

Popularity = Dynamic Choice / Popularity Rank

Choice Events

Presented (inherited from Offers)

Interested (inherited from Offers)

Added To Cart

Purchased

Choice Eligibility

None

Group Attributes

Upsell Products – Type=Product (Array) – Loading: Get Up Sell Product List (Select Product Id())

Group Eligibility

None

Dynamic Choices

Choice Id is Product Id


Table 4-42 describes the parameters for the model for the advisor Get Upsell Offers.

Table 4-42 Model for the Advisor Get Upsell Offers

Parameter Description

Model Name

Upsell Purchase Model (choice event model)

Model Setting

Use for prediction, Randomize Likelihood, Default time window, Algorithm: Bayesian

Choice Group

Upsell Offers

Base Event

Presented

Positive Events

Purchased

Partitioning Attributes

None

Excluded Attributes

None

Learn Location

Offer Response

Get Upsell Offers

Get Offers

Temporary Data Storage

None


Dependencies:

  • Informant - Initiate Session

  • Informant - Identify Customer

  • Informant - Page Turn

  • Informant - Performed Search

  • Informant - Customer Action

  • Informant - Offer Response

  • Informant - Added to Cart

  • Informant - Close Session

  • Choice Group - Offers

  • Function - Select Product Id ( )

  • Function - Get Up Sell Product List (Product Id)

4.5.1.2 Get Cross Sell Offers

The advisor Get Cross Sell Offers determines the likelihood for a customer to accept a cross sell offer.

Table 4-43 describes the parameters for the advisor Get Cross Sell Offers.

Table 4-43 Advisor Get Cross Sell Offers

Parameter Description

Advisor Name

Get Cross Sell Offers

Session Keys

Session / Current Web Interaction / Interaction Id

Request Data

Product Id (optional)

Number of Offers Rank Offers (array, optional, to be ranked, as source for dynamic choice)

External System

Web E-Commerce

Order

0

Force session close

No

Decision

Select Cross Sell Offers

Group Decision

Select Cross Sell Offers

Default Choices

None

Logic

The supplied product id is assigned to Session / Supplied Product Id. This will be used by function: Select Product Id (for Cross Sell / Up Sell). If supplied, override the default number of choices returned by decision using the Number of Offers incoming parameter.

If Rank Offer (string array) is supplied, set these into session as dynamic choice objects to be used by decision later.

Pre-condition

Base Product Id for Cross sell / Up sell must exist. Function: Select Product Id tries to find this Product Id from the following variables:

1 - Session / Supplied Product Id

2 - Session / Current Web Interaction / Current Viewed Product Id

3 - Session / Current Web Interaction / Last Added (into shopping cart) Product Id


Table 4-44 describes the parameters for the decision for the advisor Get Cross Sell Offers.

Table 4-44 Decision for Advisor Get Cross Sell Offers

Parameter Description

Decision Name

Select Cross Sell Offers

Select Choices From

Cross Sell Offers

Number of Choices to Select

5

Select at Random

No

Target Segments

Default

Priorities for Default Segment

Maximize Acceptance Likelihood 33%

Maximize Revenue 33%

Popularity 33%

Pre Selection Logic

None

Post Selection Logic

Generate learning entries for (int i = 0; i < choices.size(); i++) {choices.get(i).recordEvent("Presented");}


Table 4-45 describes the parameters for the choice group for the advisor Get Cross Sell Offers.

Table 4-45 Choice Group for Advisor Get Cross Sell Offers

Parameter Description

Choice Group Name

Cross Sell Offers

Choice Attributes

Product Id

Product Name

Product Line

Product Type

Product Category

Offer Type = "Cross Sell"

Likelihood

Scores

Maximize Acceptance Likelihood = Predicted by Cross Sell Purchase Model: Purchased

Maximize Revenue = Dynamic Choice / Unit Price

Popularity = Dynamic Choice / Popularity Rank

Choice Events

Presented (inherited from Offers)

Interested (inherited from Offers)

Added To Cart

Purchased

Choice Eligibility

None

Group Attributes

Cross Sell Products - Type=Product (Array) - Loading: Get Cross Sell Product List (Select Product Id())

Group Eligibility

None

Dynamic Choices

Choice Id is Product Id


Table 4-46 describes the parameters for the model for the advisor Get Cross Sell Offers.

Table 4-46 Model for Advisor Get Cross Sell Offers

Parameter Description

Model Name

Cross Sell Purchase Model (choice event model)

Model Setting

Use for prediction, Randomize Likelihood, Default time window, Algorithm: Bayesian

Choice Group

Cross Sell Offers

Base Event

Presented

Positive Events

Purchased

Partitioning Attributes

None

Excluded Attributes

None

Learn Location

Offer Response

Get Cross Sell Offers

Get Offers

Temporary Data Storage

None


Dependencies:

  • Informant - Initiate Session

  • Informant - Identify Customer

  • Informant - Page Turn

  • Informant - Performed Search

  • Informant - Customer Action

  • Informant - Offer Response

  • Informant - Added to Cart

  • Informant - Close Session

  • Choice Group - Offers

  • Function - Select Product Id ( )

  • Function - Get Cross Sell Product List (Product Id)

4.5.1.3 Get Promotions

The advisor Get Promotions determines the likelihood for a customer to have an interest on a presented promotion.

Table 4-47 describes the parameters for the advisor Get Promotions.

Table 4-47 Advisor Get Promotions

Parameter Description

Advisor Name

Get Promotions

Session Keys

Session / Current Web Interaction / Interaction Id

Request Data

Number of Offers

Rank Offers (array, optional, to be ranked, as source for dynamic choice)

External System

Web E-Commerce

Order

0

Force session close

No

Decision

Select Promotions

Group Decision

Select Promotions

Default Choices

None

Logic

If supplied, override the default number of choices returned by decision using the Number of Offers incoming parameter.

If Rank Offer (string array) is supplied, set these into session as dynamic choice objects to be used by decision later.

Pre-condition

None


Table 4-48 describes the parameters for the decision for the advisor Get Promotions.

Table 4-48 Decision for Advisor Promotions

Parameter Description

Decision Name

Select Promotions

Select Choices From

Promotions

Number of Choices to Select

5

Select at Random

No

Target Segments

Default

Priorities for Default Segment

Maximize Acceptance Likelihood 100%

Pre Selection Logic

None

Post Selection Logic

Generate learning entries for (int i = 0; i < choices.size(); i++) {choices.get(i).recordEvent("Presented");}


Table 4-49 describes the parameters for the choice group for the advisor Get Promotions.

Table 4-49 Choice Group for Advisor Promotions

Parameter Description

Choice Group Name

Promotions

Choice Attributes

Promotion Id

Promotion Name

Promotion Type

Promotion Period

Offer Type = "Promotion"

Likelihood

Scores

Maximize Acceptance Likelihood = Predicted by Promotion Interest Model: Interested

Choice Events

Presented (inherited from Offers)

Interested (inherited from Offers)

Choice Eligibility

None

Group Attributes

Promotion List - Type=Promotion (Array) - Data loading function: Get Promotion List ()

Group Eligibility

None

Dynamic Choices

Choice Id is Promotion Id


Table 4-50 describes the parameters for the model for the advisor Get Promotions.

Table 4-50 Model for Advisor Get Promotions

Parameter Description

Model Name

Promotion Interest Model (choice event model)

Model Setting

Use for prediction, Randomize Likelihood, Default time window, Algorithm: Bayesian

Choice Group

Promotions

Base Event

Presented

Positive Events

Interested

Partitioning Attributes

None

Excluded Attributes

None

Learn Location

Offer Response

Get Promotions

Get Offers

Temporary Data Storage

None


Dependencies:

  • Informant - Initiate Session

  • Informant - Identify Customer

  • Informant - Page Turn

  • Informant - Performed Search

  • Informant - Customer Action

  • Informant - Offer Response

  • Informant - Added to Cart

  • Informant - Close Session

  • Choice Group - Offers

4.5.1.4 Get Advertisements

The advisor Get Advertisements determines the likelihood for a customer to have an interest on a presented advertisement.

Table 4-51 describes the parameters for the advisor Get Advertisements.

Table 4-51 Advisor Get Advertisements

Parameter Description

Advisor Name

Get Advertisements

Session Keys

Session / Current Web Interaction / Interaction Id

Request Data

Number of Offers

Rank Offers (array, optional, to be ranked, as source for dynamic choice)

External System

Web E-Commerce

Order

0

Force session close

No

Decision

Select Advertisements

Group Decision

Select Advertisements

Default Choices

None

Logic

If supplied, override the default number of choices returned by decision using the Number of Offers incoming parameter.

If Rank Offer (string array) is supplied, set these into session as dynamic choice objects to be used by decision later.

Pre-condition

None


Table 4-52 describes the parameters for the decision for the advisor Get Advertisements.

Table 4-52 Decision for Advisor Get Advertisements

Parameter Description

Decision Name

Select Advertisements

Select Choices From

Advertisements

Number of Choices to Select

5

Select at Random

No

Target Segments

Default

Priorities for Default Segment

Maximize Acceptance Likelihood 100%

Pre Selection Logic

None

Post Selection Logic

Generate learning entries for (int i = 0; i < choices.size(); i++) {choices.get(i).recordEvent("Presented");}


Table 4-53 describes the parameters for the choice group for the advisor Get Advertisements.

Table 4-53 Choice Group for Advisor Get Advertisements

Parameter Description

Choice Group Name

Advertisements

Choice Attributes

Ad Id

Ad Name

Ad Type

Ad Category

Offer Type = "Ad"

Likelihood

Scores

Maximize Acceptance Likelihood = Predicted by Advertisement Interest Model: Interested

Choice Events

Presented (inherited from Offers)

Interested (inherited from Offers)

Choice Eligibility

None

Group Attributes

Ad List - Type=Ad (Array) - Data loading function: Get Ad List ()

Group Eligibility

None

Dynamic Choices

Choice Id is Ad Id


Table 4-54 describes the parameters for the model for the advisor Get Advertisements.

Table 4-54 Model for Advisor Get Advertisements

Parameter Description

Model Name

Advertisement Interest Model (choice event model)

Model Setting

Use for prediction, Randomize Likelihood, Default time window, Algorithm: Bayesian

Choice Group

Promotions

Base Event

Presented

Positive Events

Interested

Partitioning Attributes

None

Excluded Attributes

None

Learn Location

Offer Response

Get Advertisements

Get Offers

Temporary Data Storage

None


Dependencies:

  • Informant - Initiate Session

  • Informant - Identify Customer

  • Informant - Page Turn

  • Informant - Performed Search

  • Informant - Customer Action

  • Informant - Offer Response

  • Informant - Added to Cart

  • Informant - Close Session

  • Choice Group - Offers

4.5.1.5 Get Offers

The advisor Get Offers determines the likelihood for a customer to accept an offer.

The advisor Get Offers returns a mix of offers:

  • Upsell Offers

  • Cross Sell Offers

  • Promotions

  • Advertisements

Table 4-55 describes the parameters for the advisor Get Offers.

Table 4-55 Advisor Get Offers

Parameter Description

Advisor Name

Get Offers

Session Keys

Session / Current Web Interaction / Interaction Id

Request Data

Product Id (optional)

Number of Offers

Rank Offers (array, optional, to be ranked, as source for dynamic choice)

External System

Web E-Commerce

Order

0

Force session close

No

Decision

Select Offers

Group Decision

Select Offers

Default Choices

None

Logic

The supplied product id is assigned to Session / Supplied Product Id. This will be used by function: Select Product Id (for Cross Sell / Up Sell).

If supplied, override the default number of choices returned by decision using the Number of Offers incoming parameter.

If Rank Offer (string array) is supplied, set these into session as dynamic choice objects to be used by decision later.

Pre-condition

Base Product Id for Cross sell / Up sell must exist. Function: Select Product Id tries to find this Product Id from the following variables:

1 - Session / Supplied Product Id

2 - Session / Current Web Interaction / Current Viewed Product Id

3 - Session / Current Web Interaction / Last Added (into shopping cart) Product Id


Table 4-56 describes the parameters for the decision for the advisor Get Offers.

Table 4-56 Decision for Advisor Get Offers

Parameter Description

Decision Name

Select Offers

Select Choices From

Offers

Number of Choices to Select

5

Select at Random

No

Target Segments

Default

Priorities for Default Segment

Maximize Acceptance Likelihood 100%

Pre Selection Logic

None

Post Selection Logic

Generate learning entries for (int i = 0; i < choices.size(); i++) {choices.get(i).recordEvent("presented");}


Table 4-57 describes the parameters for the choice group for the advisor Get Offers.

Table 4-57 Choice Group for Advisor Get Offers

Parameter Description

Choice Group Name

Offers

Choice Attributes

Offer Type

Likelihood

Scores

Maximize Acceptance Likelihood

Choice Events

Presented

Interested

Choice Eligibility

None

Group Attributes

None

Group Eligibility

None

Dynamic Choices

None


Dependencies:

  • Informant - Initiate Session

  • Informant - Identify Customer

  • Informant - Page Turn

  • Informant - Performed Search

  • Informant - Customer Action

  • Informant - Offer Response

  • Informant - Added to Cart

  • Informant - Close Session

  • Choice Group - Upsell Offers

  • Choice Group - Cross Sell Offers

  • Choice Group - Promotions

  • Choice Group - Advertisements

4.5.2 Analysis Advisors

This section describes the following advisors:

4.5.2.1 Get Abandonment Propensity

The advisor Get Abandonment Propensity returns the likelihood of abandonment for a customer.

Table 4-58 describes the parameters for the advisor Get Abandonment Propensity.

Table 4-58 Advisor Get Abandonment Propensity

Parameter Description

Advisor Name

Get Abandonment Propensity

Session Keys

Session / Current Web Interaction / Interaction Id

Request Data

None

External System

Web E-Commerce

Order

0

Force session close

No

Decision

Select Abandonment Propensity

Group Decision

Select Abandonment Propensity

Default Choices

None

Logic

None

Pre-condition

None


Table 4-59 describes the parameters for the decision for the advisor Get Abandonment Propensity.

Table 4-59 Decision for Advisor Get Abandonment Propensity

Parameter Description

Decision Name

Select Abandonment Propensity

Select Choices From

Abandonment

Number of Choices to Select

1

Select at Random

No

Target Segments

Default

Priorities for Default Segment

Maximize Likelihood 100%

Pre Selection Logic

None

Post Selection Logic

None


Table 4-60 describes the parameters for the choice group for the advisor Get Abandonment Propensity.

Table 4-60 Choice Group for Advisor Get Abandonment Propensity

Parameter Description

Choice Group Name

Abandonment

Choice Attributes

Name

Likelihood = Get Choice Likelihood ("AbandonmentModel", this)

Scores

Maximize Likelihood = Likelihood

Choice Events

None

Choice Eligibility

None

Group Attributes

None

Group Eligibility

None

Dynamic Choices

None


Table 4-61 describes the parameters for the model for the advisor Get Abandonment Propensity.

Table 4-61 Model for Advisor Get Abandonment Propensity

Parameter Description

Model Name

Abandonment Model (choice model)

Model Setting

Use for prediction, Randomize Likelihood, Default time window, Algorithm: Bayesian

Choice Group

Abandonment

Mutually Exclusive

Yes

Partitioning Attributes

None

Excluded Attributes

None

Learn Location

On session close

Temporary Data Storage

None

Model Name

Abandonment Model (choice model)


Dependencies:

  • Informant - Close Session

  • Function - Get Choice Likelihood (Model Name, Choice)

4.5.2.2 Get Likely Web Action

The advisor Get Likely Web Action predicts the most likely Web Action that a particular customer will perform next.

Table 4-62 describes the parameters for the advisor Get Likely Web Action.

Table 4-62 Advisor Get Likely Web Action

Parameter Description

Advisor Name

Get Likely Web Action

Session Keys

Session / Current Web Interaction / Interaction Id

Request Data

None

External System

Web E-Commerce

Order

0

Force session close

No

Decision

Select Likely Web Action

Group Decision

Select Likely Web Action

Default Choices

None

Logic

None

Pre-condition

None


Table 4-63 describes the parameters for the decision for the advisor Get Likely Web Action.

Table 4-63 Decision for Advisor Get Likely Web Action

Parameter Description

Decision Name

Select Likely Web Action

Select Choices From

Web Actions

Number of Choices to Select

1

Select at Random

No

Target Segments

Default

Priorities for Default Segment

Maximize Likelihood 100%

Pre Selection Logic

None

Post Selection Logic

None


Table 4-64 describes the parameters for the choice group for the advisor Get Likely Web Action.

Table 4-64 Choice Group for Advisor Get Likely Web Action

Parameter Description

Choice Group Name

Web Actions

Choice Attributes

Name

Likelihood = Get Choice Likelihood ("WebActionModel", this)

Scores

Maximize Likelihood = Likelihood

Choice Events

None

Choice Eligibility

None

Group Attributes

None

Group Eligibility

None

Dynamic Choices

None


Table 4-65 describes the parameters for the model for the advisor Get Likely Web Action.

Table 4-65 Model for Advisor Get Likely Web Action

Parameter Description

Model Name

Web Action Model (choice model)

Model Setting

Use for prediction, Randomize Likelihood, Default time window, Algorithm: Bayesian

Choice Group

Web Actions

Mutually Exclusive

No

Partitioning Attributes

None

Excluded Attributes

None

Learn Location

Customer Action

Temporary Data Storage

None

Model Name

Web Action Model (choice model)


Dependencies:

  • Informant - Customer Action

  • Function - Get Choice Likelihood (Model Name, Choice)

4.5.2.3 Get Likely Web Duration

The advisor Get Likely Web Duration predicts the length of time that a particular customer will spend on the site.

Table 4-66 describes the parameters for the advisor Get Likely Web Duration.

Table 4-66 Advisor Get Likely Web Duration

Parameter Description

Advisor Name

Get Likely Web Duration

Session Keys

Session / Current Web Interaction / Interaction Id

Request Data

None

External System

Web E-Commerce

Order

0

Force session close

No

Decision

Select Likely Web Duration

Group Decision

Select Likely Web Duration

Default Choices

None

Logic

None

Pre-condition

None


Table 4-67 describes the parameters for the decision for the advisor Get Likely Web Duration.

Table 4-67 Decision for Advisor Get Likely Web Duration

Parameter Description

Decision Name

Select Likely Web Duration

Select Choices From

Web Site Duration

Number of Choices to Select

1

Select at Random

No

Target Segments

Default

Priorities for Default Segment

Maximize Likelihood 100%

Pre Selection Logic

None

Post Selection Logic

None


Table 4-68 describes the parameters for the choice group for the advisor Get Likely Web Duration.

Table 4-68 Choice Group for Advisor Get Likely Web Duration

Parameter Description

Choice Group Name

Web Site Duration

Choice Attributes

Name

Likelihood = Get Choice Likelihood ("WebSiteDurationModel", this)

Scores

Maximize Likelihood = Likelihood

Choice Events

None

Choice Eligibility

None

Group Attributes

None

Group Eligibility

None

Dynamic Choices

None


Table 4-69 describes the parameters for the model for the advisor Get Likely Web Duration.

Table 4-69 Model for Advisor Get Likely Web Duration

Parameter Description

Model Name

Web Site Duration Model (choice model)

Model Setting

Use for prediction, Randomize Likelihood, Default time window, Algorithm: Bayesian

Choice Group

Web Site Duration

Mutually Exclusive

Yes

Partitioning Attributes

None

Excluded Attributes

None

Learn Location

On session close

Temporary Data Storage

None

Model Name

Web Site Duration Model (choice model)


Dependencies:

  • Informant - Customer Action

  • Function - Get Choice Likelihood (Model Name, Choice)

4.5.2.4 Get Web Support Types

The advisor Get Web Support Types returns the likelihood for a customer to accept a chat request or email support if offered.

Table 4-70 describes the parameters for the advisor Get Web Support Types.

Table 4-70 Advisor Get Web Support Types

Parameter Description

Advisor Name

Get Web Support Types

Session Keys

Session / Current Web Interaction / Interaction Id

Request Data

None

External System

Web E-Commerce

Order

0

Force session close

No

Decision

Select Web Supports

Group Decision

Select Web Supports

Default Choices

None

Logic

None

Pre-condition

None


Table 4-71 describes the parameters for the decision for the advisor Get Web Support Types.

Table 4-71 Decision for Advisor Get Web Support Types

Parameter Description

Decision Name

Select Web Support Types

Select Choices From

Web Support Types

Number of Choices to Select

2

Select at Random

No

Target Segments

Default

Priorities for Default Segment

Maximize Likelihood 100%

Pre Selection Logic

None

Post Selection Logic

Generate learning entries for (int i = 0; i < choices.size(); i++) {choices.get(i).recordEvent("presented");}


Table 4-72 describes the parameters for the choice group for the advisor Get Web Support Types.

Table 4-72 Choice Group for Advisor Get Web Support Types

Parameter Description

Choice Group Name

Web Support Types

Choice Attributes

Name

Likelihood of Usage = Predicted by Web Support Usage Model: Used

Threshold = 0.5

Scores

Maximize Likelihood = Likelihood of Usage

Choice Events

Presented

Used

Choice Eligibility

Choice / Likelihood of Usage > choice / Threshold

Group Attributes

None

Group Eligibility

None

Dynamic Choices

None


Table 4-73 describes the parameters for the model for the advisor Get Web Support Types.

Table 4-73 Model for Advisor Get Web Support Types

Parameter Description

Model Name

Web Support Usage Model (choice event model)

Model Setting

Use for prediction, Randomize Likelihood, Default time window, Algorithm: Bayesian

Choice Group

Web Support Types

Base Event

Presented

Positive Events

Used

Partitioning Attributes

None

Excluded Attributes

None

Learn Location

On session close

Temporary Data Storage

None


Dependencies:

  • Informant - Web Support Feedback

4.6 Functions

This section describes the functions used in the RTD_Base_ECommerce Inline Service.

Table 4-74 RTD_Base_ECommerce Functions

Function Inputs Outputs Area Utilized In Comments

Get Ad List

None

Advertisements (Array of Ad)

Used for Advertisements dynamic choices.

See corresponding functions: Set Session Rank Offers.

This function is used to return array of Ad entity for Advertisement dynamic choice. The result is either ranked advertisements array supplied as incoming parameter of advisors or array of advertisements loaded using data source mapping.

Get Agent Interaction Reasons In Past Days

Past Agent Interactions (Array of Agent Interaction),

Days (Integer)

Interaction Reasons (Array of String)

Used for deriving values for Interaction History entity.

This function returns array of interaction reason in past given days.

Get Agent Interaction Types In Past Days

Past Agent Interactions (Array of Agent Interaction),

Days (Integer)

Interaction Types (Array of String)

Used for deriving values for Interaction History entity.

This function returns array of interaction types in past given days.

Get Campaign Names In Past Days

Campaign Items (Array of Campaign Item), Days (Integer)

Campaign Names (Array of String)

Used for deriving values for Campaign History entity

This function returns array of campaign names delivered in past given days.

Get Choice Likelihood

Model Name (String), Choice (Choice)

Likelihood (Double)

Used for scoring in Abandonment, Web Actions, Web Site Duration choice groups

This function returns likelihood for a given inputted choice that is part of a choice model (as opposed to a choice event model). As inputs, the user must pass in a choice and the model name that choice is part of.

Get Cross Sell Product List

Product Id (String) to be used as base product for cross sell

Cross Sell Products (Array of Product)

Used for Cross Sell dynamic choices.

See corresponding functions: Set Session Rank Offers.

This function is used to return array of Product entity for Cross Sell dynamic choice. The result is either ranked products array supplied as incoming parameter of advisors or array of cross sell products loaded using data source mapping.

Get Customer Type

Person (Person), Organization (Organization)

Customer Type (String)

Used by Customer entity.

This function is used to determine customer type.

Get Days Left

Expiration Date (Date)

Number of Days Left (Integer)

Used by Promotion entity.

This function calculates the number of days from now to expiryDate.

Get Days Since Last Agent Interaction

Past Agent Interactions (Array of Agent Interaction)

Number of Days (Integer)

Used by Interaction History entity.

This function calculates the number of days from last agent interaction to now.

Get Days Since Last Campaign

Campaign Items (Array of Campaign Item)

Number of Days (Integer)

Used by Campaign History entity.

This function calculates the number of days from last campaign to now.

Get Days Since Last Purchase

Purchased Items (Array of Purchased Item)

Number of Days (Integer)

Used by Purchase History entity.

This function calculates the number of days from last purchase to now.

Get Days Since Last Web Interaction

Past Web Interactions (Array of Web Interaction)

Number of Days (Integer)

Used by Interaction History entity.

This function calculates the number of days from last web interaction to now.

Get Duration In Days

Start Date (Date), End Date (Date)

Number of Days (Integer)

Used by Promotion entity and various other functions.

This function calculates the number of days from start date to end date.

Get Duration In Minutes

Start Time (Date), End Time (Date)

Number of Minutes (Integer)

NA.

This function calculates the number of minutes from start time to end time.

Get Gender

Male (Boolean)

Gender In Text (String)

Used by Customer entity.

This function determines the gender of a person. This converts the given gender type from boolean to string. In Male case, genderMale is boolean value=true fetched from data source.

Get Interaction Types In Past Days

Past Agent Interactions (Array of Agent Interaction), Past Web Interactions (Array of Web Interaction),

Days (Integer)

Interaction Type in Text (String)

Used by Interaction History entity.

This function returns interaction type in past given days.

Get Last Agent Interaction Status

Past Agent Interactions (Array of Agent Interaction)

Last Status (String)

Used by Interaction History entity

This function returns last agent interaction status.

Get Last Agent Interaction Type

Past Agent Interactions (Array of Agent Interaction)

Last Type (String)

Used by Interaction History entity

This function returns last agent interaction type.

Get Last Campaign Category

Campaign Items (Array of Campaign Item)

Category (String)

Used by Campaign History entity.

This function returns last campaign category.

Get Last Campaign Delivery Method

Campaign Items (Array of Campaign Item)

Delivery Method (String)

Used by Campaign History entity.

This function returns last campaign delivery method.

Get Last Campaign Name

Campaign Items (Array of Campaign Item)

Name (String)

Used by Campaign History entity.

This function returns last campaign name.

Get Last Campaign Type

Campaign Items (Array of Campaign Item)

Type (String)

Used by Campaign History entity.

This function returns last campaign type.

Get Last Interaction Type

Days Since Last Agent Interaction (Integer), Days Since Last Web Interaction (Integer)

Interaction Type (String)

Used by Interaction History entity.

This function returns last interaction type. Return value will be Agent, Web, or Both.

Get Last Purchase Amount

Purchased Items (Array of Purchased Item)

Purchase Amount (Double)

Used by Purchase History entity.

This function returns last purchase amount.

Get Last Purchased Product

Purchased Items (Array of Purchased Item)

Product Name (String)

Used by Purchase History entity.

This function returns last purchased product name.

Get Last Purchased Product Line

Purchased Items (Array of Purchased Item)

Product Line (String)

Used by Purchase History entity.

This function returns last purchased product line.

Get Number Of Agent Interaction In Past Days

Past Agent Interactions (Array of Agent Interaction),

Days (Integer)

Count (Integer)

Used by Interaction History entity.

This function counts number of agent interactions in past given days.

Get Number Of Web Interaction In Past Days

Past Web Interactions (Array of Web Interaction),

Days (Integer)

Count (Integer)

Used by Interaction History entity.

This function counts number of web interactions in past given days.

Get Performed Web Action Names In Past Days

Past Web Interactions (Array of Web Interaction),

Days (Integer)

Action Names (Array of String)

Used by Interaction History entity.

This function returns array of web action names performed in past given days.

Get Product Lines Owned

Purchased Items (Array of Purchased Item)

Product Lines (Array of String)

Used by Purchase History entity.

This function returns product lines of customer owned.

Get Promotion List

None

Promotions (Array of Promotion)

Used for Promotion dynamic choices.

See corresponding functions: Set Session Rank Offers.

This function is used to return array of Promotion entity for Promotion dynamic choice. The result is either ranked promotion array supplied as incoming parameter of advisors or array of promotions loaded using data source mapping.

Get Specific Choice Likelihood

Model Name (String), Choice Name (String)

Likelihood (Double)

None.

See similar function: Get Choice Likelihood

This function returns likelihood for a given inputted choice that is part of a choice model (as opposed to a choice event model). As inputs, the user must pass in the value for the Name attribute assigned to the choice and the model name that choice is part of.

Get Total Amount Spent

Purchased Items (Array of Purchased Item)

Amount (Double)

Used by Purchase History entity.

This function sums up the amount of customer spent so far.

Get Total Amount Spent in Last 90 Days

Purchased Items (Array of Purchased Item)

Amount (Double)

Used by Purchase History entity.

This function sums up the amount of customer spent in past 90 days.

Get Total Duration In Minutes

Start Time (Date)

Number of Minutes (Integer)

Used by Web Interaction entity.

This function calculates number of minutes between the given start time and now.

Get Up Sell Product List

Product Id (String) to be used as base product for up sell

Up Sell Products (Array of Product)

Used for Up Sell dynamic choices.

See corresponding functions: Set Session Rank Offers.

This function is used to return array of Product entity for Up Sell dynamic choice. The result is either ranked products array supplied as incoming parameter of advisors or array of up sell products loaded using data source mapping.

Get Year

Date (Date)

Year (Integer)

Used by Customer entity.

This function returns the year part of Date.

Is Web Support Type Eligible

Web Support Types Choice (Web Support Types Choice)

Eligible (Boolean)

Used by Web Support Types choice group.

This function returns the eligibility of a Web Support Types Choice.

Maximum

X (Integer), Y (Integer)

Maximum (Integer)

Used by Interaction History entity.

This function returns the greater parameter.

Minutes To Now

Start Time (Date)

Minutes (Integer)

Used by Total Duration In Minutes function

This function returns time between given time in the past and now.

Multiply

a (Double, b (Double)

Result (Double)

Used by various RTD elements

This function multiplies the given parameters.

Property Reflect

NA

NA

NA

This function is used for testing. This dumps the values of session attributes.

Select Product Id

None

Product Id (String)

Used as base product for Up Sell and Cross Sell dynamic choices.

This function determines what Product Id to be used as base Product Id for Up Sell or Cross Sell.

Set Choice Event Model

Choice Event Model Name (String),

Choice Name (String), Choice Event (String)

None

Used by Web Support Feedback informant.

This function is used to store all of the logic needed to set the choice models tied to the choice groups.

Set Choice Model

Choice Model Name (String),

Choice Name (String)

None

Used by Customer Action and Close Session informants.

This function is used to store all of the logic needed to set the choice models tied to the choice groups.

Set Session Rank Offers

Offers (Array of String)

None

Used by all offer-oriented advisors to set the optionally supplied string array of keys into session for later use as dynamic choice source.

See corresponding functions: Get Ad List, Get Promotion List, Get Cross Sell Product List, Get Up Sell Product List.

This procedure is used to parse and set Rank Offers before decision selection.

Years To Now

Start Time (Date)

Years (Integer)

Used by Customer entity.

This function calculates the number of years from given date to now.