Conventions March 22nd 2002

1
ConventionsMarch 22nd 2002
2
Discovery Conventions

• Discovery Conventions - positioning
• A holistic approach to KID and the generic
  attribute set
• The Contextual Attributes and Product Types
• Relationship to complex topic clustering and
  usage algorithms
• Coverage to date

3
Conventions - Positioning
Naming Standards
Metadata Standards
Exchange Template Standards
Presentation Format/Layout Standards
Technology Standards
Naming Standards
Metadata Standards
Exchange Template Standards
Presentation Format/Layout Standards
Technology Standards
Lots of different types of standards
4
Conventions - Positioning
Naming Standards
Metadata Standards
Exchange Template Standards
Presentation Format/Layout Standards
Technology Standards
Naming Standards
Metadata Standards
Exchange Template Standards
Presentation Format/Layout Standards
Technology Standards
Focus on naming and metadata
5
Conventions - Positioning
Naming Standards
Metadata
Asset names (e.g. wells, fields, surveys)
Etc
Product Types Buckets
Generalist metadata
Dti Well Name
Specialist Filing Names
Specialist metadata and schema
Aimed at the high level finding and discovery
attributes
6
Conventions - Positioning
People/ Roles
Processes
Tools
Work Products
Have Roles/ Disciplines
Execute Processes
Using Tools
To produce work products
Applied holistically to Roles, Processes, Tools
and Work Products
7
Conventions - Positioning
Indexed at the catalogue layer for
integration and sharing
Product Type Buckets
Generalist metadata
Discovery Attribute Set How you find
things e.g. author, source, contextual
attributes etc.
e.g. Well Header
Specialist Filing Names
Specialist metadata and schema
Held in underlying specialist systems
Typical datamodel e.g. class, status, location
etc.
e.g. Well Header.
POSC HdrML
Sometimes there is a 1-to-1 relationship between
generalist buckets and detailed filing
nomenclature.
Technical Example - Well Header (Typical Database
View)
8
Conventions - Positioning
Indexed at the catalogue layer for
integration and sharing
Product Type Buckets
Generalist metadata
Discovery Attribute Set How you find
things e.g. author, source, contextual
attributes etc.
e.g. Joined Cased Hole Well Log Data
e.g. Gamma Density Neutron Sonic etc.
Specialist Filing Names
Specialist metadata and schema
Held in underlying specialist systems
Detailed item specific metadata e.g. logging run,
tool type, top, base etc.
POSC PWLS
CDA Hdcpy
Sometimes there is a 1-to-many match between
generalist buckets and detailed filing
nomenclature
Technical Example - Well Log Data
9
Taking a holistic approach to KID
The Knowledge, Information and Data (KID)
logs
lt
People
Workflow
Database Views
logs
Sections
Tools
lt
Maps
logs
Database Views
People
logs
Workflow

A generic set of attributes and contextual
valid-values were used to describe any container
of knowledge, information or data
10
Discovery Generic Attribute Set
Control
Bibliographic
Title
Content Owner
Retention
Usage
Author
AFE
Description
Unique Ref
Review Date
Recipient
Published Status
Approved By
Media
Logical Fomat
Source Organisation.
Tag/EDCC
Cross References
Security Entitlement
Revision Code
Published Date
Language
Location
Representation
Coverage
Relationships
Datum
Compound Document
Geographical Coords/poly
Collection
Scale
Contextual
System
Asset UID
Creator Discipline
Asset Type
File Size
Created By
Business Process
Product Type
Product Group
Cust. Discipline
Last Modified
KID Type
A range of attributes to enable items to be
effectively discovered and evaluated
11
Discovery Generic Attribute Set
Control
Bibliographic
Title
Content Owner
Retention
Usage
Author
AFE
Description
Unique Ref
Review Date
Recipient
Published Status
Approved By
Media
Logical Fomat
Source Organisation.
Tag/EDCC
Cross References
Security Entitlement
Revision Code
Published Date
Language
Location
Representation
Coverage
Relationships
Datum
Compound Document
Geographical Coords/poly
Collection
Scale
Contextual
System
Asset UID
Creator Discipline
Asset Type
File Size
Created By
Business Process
Product Type
Product Group
Cust. Discipline
Last Modified
KID Type
Focusing on the contextual attributes which
encompass the common ways in which people search
for knowledge, information and data
12
Discovery Contextual Attributes
Customer Discipline
Product Group
Key search contexts 1. Business process of
creation 2. Creator Discipline 3. Customer
Discipline 4. KID Type 5. Product Group
KID Type
Creator Discipline
Business Process of creation
Different unstructured keywords have been used
historically, implemented differently within
isolated systems to manage content by media or
topic
13
Creating Product Type Names
e.g. Raw data/as-built records Lesson
Learnt Plans, Programmes Proposals,
applications Contracts and agreements Operations
Report Correspondence Meeting Minutes Audits and
Inspections Analysis and Interpretation etc.
e.g. Well Proposal Well Rock/Core Data Well
Completion Programme
e.g. Well Drilling Well Data Acquisition Determine
HC Structure Seismic Interpretation Evaluate
rock/fluid properties Petrophysical
Interpretation Calculate hydrocarbons in
place etc.
e.g. WellRock/Core WellEquipmentTubing WellHea
der WellSynthetics WellStrat Pick FieldReserves
SurveySite Survey PipelineUllage etc.
e.g. Geology Geophysics Petrophysics Reservoir
Engineering Production Technology etc.
Product Types apply to data in databases and
documents
14
Example
Sample Product Types and modelled contexts
15
Relationship to complex topic clustering

• The Discovery Standard is a piece of intellectual
  property. It is the codification of common EP
  user contexts which can be used to teach
  clustering algorithms.
• Complex search engines are able to create
  clusters of ad-hoc topics below the level of
  granularity of the business vocabulary using an
  EP industry thesaurus. These clusters are
  static but get added to when more products are
  put into the corporate memory and as the
  thesaurus is improved. e.g.
• Clastic Reservoirs
• Horizontal Drilling
• Depth Imaging beneath Salt
• Knowledge networks can also create clusters (e.g.
  neural network/statistical) of ad-hoc topics by
  monitoring usage patterns - so called creation
  of new linkages. These clusters are dynamic, as
  they change with usage patterns.

Discovery standards, clustering algorithms and
usage patterns compliment, not replace one another
16
Current Scope of Discovery Conventions

• Subsurface work products (approx 800 Product
  Types and modelled contexts created)
• Wells - First Draft Complete
• Field/Prospect/Reservoir - Pre-Draft Complete
• License - Needs More Work
• Survey - Only preliminary seabed survey product
  types (build on DEAL metadata?)
• Topsides and above ground work products
• Facilities/Platforms, pipelines No work to
  date (OFS Portal/Pidex/POSC Caesar?)
• Corporate/organisational work products
• Finance, HR, Commercial, Tax, IT, Strategy,
  Corporate Affairs etc. No work to date
• Governance structures
• Pre-draft of Subsurface Information Management
  governance
• role definition, assignment of responsibilities
• best practices
• KID standards
• appplications and databases

The main thrust of Discovery Conventions has been
on the multidisciplinary wellfile work products
and Subsurface IM Best Practices Roles