apml-1.0-draft-1 - APML.Public.General

Abstract

Attention Profiling Markup Language (APML) provides mechanisms for exchanging and storing attention data across different applications. This specification provides the base features to be supported by 1.0 compliant implementations, and is intended to be supplemented by a secondary specification detailing mechanisms for\ optional extensions.

APML defines two broad category of data, being Explicit Data and Implicit Data. Explicit Data is data gathered directly from users, whilst Implicit Data is gained through more programmatic analysis strategies, as available in individual applications.

Within these broad categories, there are four basic types of elements, being Concept, Sources, People and Locations.

1. Notation and Conventions
2. Definitions
3. Overview
4. Document Head
5. Document Body
    5.1. Profiles
    5.2. Implicit Data
    5.3. Explicit Data
    5.4. Concepts
    5.5. Sources
    5.6. People
    5.7. Locations
    5.8. Element Values
Appendix A. Appendix A - APML Schema Definition
6. References
§ Authors' Addresses

TOC

1. Notation and Conventions

The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in [RFC2119] (Bradner, B., “Key words for use in RFCs to Indicate Requirement Levels,” .). Domain name examples use [RFC2606] (Eastlake, D. and A. Panitz, “Reserved Top Level DNS Names,” .).

TOC

2. Definitions

target
Used to refer to the entity that APML is describing. Whilst typically synonymous with a user, APML may also be used to describe the content of an internet resource, such as a webpage or feed.

TOC

3. Overview

An APML document is arranged into a hierarchy consisting of three levels. At the top level, target information is grouped into a series of Profiles. Within each of the Profiles, to categories of information are available - implicit and explicit data. Within Implicit and Explicit Data, four different data types are maintained - Concept, Sources, People and Locations.

TOC

4. Document Head

Each APML document contains a standard pre-amble that provides meta-data about the document and it's target.

The head element will typically be presented in a form similar to:

<Head> <Title>My APML</Title> <Generator>http://www.example.com/apml_generator</Generator> <Target type="email">user@example.com</Target> </Head>

Configuration:

Title
The Title of the APML document. The value has no specific purpose other than for presentation.
Generator
URI for the application generating the APML document.
Target
The Target of the APML document. May be in many different forms, as defined by the type attribute. This specification initially defines three different Target types: email - the email of the user; emailhash - an MD5 hash of the user's email; url - URL of an Internet based resource

TOC

5. Document Body

The main content of an APML document is contained with the Body element. The following sections detail the elements within the body.

TOC

5.1. Profiles

From the outset, APML aims to acknowledge that users will have different aspects of their interests. Typical profile might be Home or Work, therefore allowing any application working with the APML to participate in the segmentation.

A profile element will typically be presented in a form similar to:

Configuration:

name
Provides an identifier for the profile, allowing it to be referenced both externally and internally.

TOC

5.2. Implicit Data

Implicit Data represents data that has been implicitly gathered about the target. This may be through programmatic analysis, or other techniques. An Implicit Data node may contain Concept, Source, Person and Location nodes. Any child node of an Implicit Data block MUST be annotated with a from and modified attribute pair, in order to ensure seamless interaction and operability between multiple contributing applications within a single APML profile.

An ImplicitData element will typically be presented in a form similar to:

TOC

5.3. Explicit Data

Explicit Data represents data that has been explicitly gathered from the target. As Explicit Data is shared between all applications, the target should be directly involved in the manipulation, as no conflict resolution procedures are available.

An ExplicitData element will typically be presented in a form similar to:

TOC

5.4. Concepts

A concept is a textual entity, usually either a work or series of words that represents a point of interest for the target. At the most basic level, a concept MUST contain both a name and an assigned value. If the application has further information available about the concept, such as a URI describing it, then this can also be provided.

The Concepts element consists of a series of Concept elements, typically presented in a form similar to:

Configuration:

key
The key of the concept. Generally the plaintext key for the concept.
value
The numeric value assigned to the relevance of the concept. Between -1.0 and 1.0. Further defined in the Element Values section.
rdf:about
URI providing further detail about the concept that can be further explored. Optional, as many applications may not have this information

TOC

5.5. Sources

A source details the relevancy of a given content source to the target. This specification initially defines mechanisms for describing a URI addressable source, however, later extensions may detail other possible types of entity. A Source may optionally contain a series of Author elements, that provide details about the relevance of various Source authors.

The Sources element consists of a series of Source elements, typically presented in a form similar to:

Source Configuration:

key
The URI of the source
value
The numeric value assigned to the relevance of the source. Between -1.0 and 1.0. Further defined in the Element Values section.
type
A mime-type for the source.

Author Configuration:

key
The identifier of the author. Valid values are dependent on the type of the source.
value
The numeric value assigned to the relevance of the author. Between -1.0 and 1.0. Further defined in the Element Values section.

TOC

5.6. People

A person element provides details of a person relevant to the target's APML. Given that the actual specific identity of the person is not always relevant to the calculation of the APML, and also likely to introduce significant identity issues, a Person element focusses on identifying the information through a URI to an external APML file.

The People element consists of a series of Person elements, typically presented in a form similar to:

Person Configuration:

key
A unique identifier for the person. This specification does not detail the format of the key element.
value
The numeric value assigned to the relevance of the Person. Between -1.0 and 1.0. Further defined in the Element Values section.
link
The link to the APML file containing details of the target Person.

TOC

5.7. Locations

A location element provides details of a location relevant to the target.

The Locations element consists of a series of Location elements, typically presented in a form similar to:

Location Configuration:

key
A unique identifier for the location.
value
The numeric value assigned to the relevance of the Location. Between -1.0 and 1.0. Further defined in the Element Values section.
lat
The latitude of the location.
long
The longitude of the location.

TOC

5.8. Element Values

Every data element in APML is annotated with a value element. A value is specified between -1.0 and 1.0. Whilst this specification does not provide definitive details as to the use of these numbers, the following is made available as a guide:

Negative values indicate specific dis-interest in a topic.

Positive values indicate specific interest in a topic

A zero-value indicates neither interest nor disinterest, and should be considered the default value for any non-listed term in APML

Note that the lack of numeric clarity in this specification does not preclude an extension from later providing more detailed definitions of values.

Version:

SpgmvSM.html	Oct 6
XFjKrUmW.html	Sep 16
services-supporting-apml	Sep 3
apml-1.0-draft-1	Aug 8
apml-lite	Jul 31
services-indicating-intended-support	Jul 31
adoption	Jul 31
apml-json	Jul 31
extensions	Jul 31
apml-1-0---proposed	Jul 31

Abstract

Table of Contents

1. Notation and Conventions

2. Definitions

3. Overview

4. Document Head

5. Document Body

5.1. Profiles

5.2. Implicit Data

5.3. Explicit Data

5.4. Concepts

5.5. Sources

5.6. People

5.7. Locations

5.8. Element Values