SharePoint Saturday New york City - The importance of metadata #spsnyc

THE IMPORTANCE OF METADATA

Vincent BIRET

7/25/15

ABOUT MEVincent BIRET

SharePoint MVP

Products Team Tech Lead

Negotium

THANKS

GOALS

•Why are we going to talk about metadata?• Improve navigation

• Improve findability(fact that we find what we’re looking for)

• Improve discoverability(fact that we find what we’re not looking for)

• Improve user experience

• Allow to build a governance strategy

• Save money/add value for the company

AGENDA

•Theory

•SharePoint

•Vision

•Graph

•Conclusion

REMINDERSBack to school!

« Metadata » very fashinoable term but what does it mean?

Defintion, examples and standards

WHAT ARE METADATA?

METADATA ARE DEFINED BY THEIR FUNCTION

• Ease search/discovery of information• Describe content and

relationships between files

• Classify content based on targeted audience

• Help the SEO of a page

• Ease the interoperability• Share information

• Ease management and archiving• Give an information about

the document’s lifecycle

• Manage sets of resources

• Manage electronic archives

•Manage rights• Copyright

• Security

METADATA TYPES

• Bibliographic Metadata: allowing to access the document (author, title, creation date, modification date…)

• Administrative Metadata: legal, lifecycle, copyright, processing

• Technical Metadata: format, how it was produced, storage.

• Description Metadata: to understand the content

DUBLIN CORE

Dublin core elements

Title Given name of the resourceAuthor Entity mainly responsible for resource’s content creation.

Subject Subject of the content (controlled vocabulary and classification schemas are suggested)

Description Description of the content (eg. Summary, table of content, or simply text…)

Editor Entity responsible for resource’s diffusion (company, university…)

Contributor Entity which contributed to the creation of the content.

Date Date associated to an event in resource’s lifecycle (format : see ISO 8601)

TypeNature ou genre (ex. categories, functions, general kind… controlled vocabulary recommended, see Dublin Core)

Format Physical or digital materialization of the content (mime-type)

Identifier Disambiguated (unique) reference to the content (examples : URI, URL, DOI, ISBN)

Source Reference to a resource from which this one has been derived.

Language Content’s language

Relation Reference to another related resource.

Coverage Range or coverage of the resource in space or time or legally.

Rights Security and copyright information.

METADATA STRUCTUREHow to define/organize it

CONTROLLED VOCABULARY

• A controlled vocabulary is a glossary designed to allow knowledge organization in order to optimize information lookup/search.

•Classification schemes (taxonomy, thesaurus, ontology) use a controlled vocabulary.

TAXONOMY

• Subsumption relationship “is-a”

• Transitivity

• If C is a B and B is an A, C is an AVehicules

Land Vehicules

Car

Truck

Bus

Air Vehicules

Airplane

Balloon

Rocket

Sea Vehicule

Boat

Submarine

THESAURUS

• Taxonomy++

• Hierarchy + propertiesConept or potential term Car

Generic Term Véhicule

Specific Term  

Domains vehicule - transport [MT 3330]

Regional variations  

French equivalent voiture

Associated Terms [TA] craft : transport [MT 6005]automobile technic [MT 3510]

ONTOLOGY

Partitive relationships: Part-whole (car, chassis, wheels, engine) Composite-component (steel, iron)Content-container (tools, toolbox) Membre-collection( football player, team).

Associative relationships :Cause-effet (detonator, explosion)Agent-action-result (drawer, drawing) Producer-product (waste site, recycled product) Tool-tool (scewdriver, screw) Topologic relationships (oasis, desert).

ONTOLOGY OF A CAR

FOLKSONOMY

• Comes from “folk” (people) and “taxonomy”. Indexing spontaneously by keywords without pre-defined classification.

• This process is collaborative. (sometimes also called social tagging)

FACETS

•Description of a resource following multiple axes

•Allows you to search/refine by facet

CONCEPTS

• Term = denomination + concept

• Denomination = linguistic expression

• Concept = meaning

• The term has to have a context to understand the concept

• Ex table• Coffee table?

• Table of content?

• Timetable?

• Discussion table?

NAMED ENTITIES

•Main types of entities•Person’s names

•Places

•Organizations

•Products

•Dates

•Currencies

RECAP

Schema Taxonomy Thesaurus Ontology Folksonomy Facets

Organization

Simple hierarchical relationships

Hierarchical relationships (hyponymy, hypernymy) , associative relationships and equivalent relationships

Concepts with types, properties and relationships

Tags set by users as the like

Allows to describe a resource following multiple azes

Relationship type

« is a » « is a »« sort of »« related to »« synonym »« equivalent »

« contains »« is located»« uses »« produces »« … »

  « price »« product »« kind »

Context   Library Web 3.0 Web 2.0  Example Dewey Rameau Cyc, Wordnet Flickr, YouTube Amazon

NOT SLEEPING YET?Good let’s talk about SharePoint now

SHAREPOINTWhat can it do for you?

FOLDERS

•A folder structure is already a form of metadata

•Very simple to set up

•Complicated to change

•Trick•Don’t use it unless you have permissions related needs

TAXONOMY

•Advantages:• Search

• Filtering

• Advanced control

• Consistency

•Disadvantages:• Takes time to set up

•Tips:•Use most issued search requests to improve it!

•Delegate management of some branches

• If you don’t have a clear idea yet, use folksonomy

ENTERPRISE KEYWORDS

•Advantages:• Easier fro end users

• Takes into consideration keywords inside of documents

•Disadvantages:•Harder to manage

•No context during indexation

FOLKSONOMY

Content « like »Tagging (with user’s words)

NotesUser Profile

Filtered NavigationSearch refiner

Topics and power users to followDeprecated on SharePoint online!!!

DOCUMENTS SETS

« this document is related to that other one »Creates a documents « group » modelProperties can be common or individual

Trick:Can also be used for security purposes (not primary role)Don’t hesitate to create new content types inhereting from it

COLUMNS AND CONTENT TYPES

• Information about what the content is (CT)

• Adds properties (columns)

• Allows filtering, refinement, search….

• Allows to have multiple entry forms(CT)

NAVIGATION BY TAXONOMY

• Enhances nvigation

• Prepares users to use refiners

• Easy to set up

• Is a great alternative to/transition from folders

WIKIS

• Quick pages creation system

• Allows to link web content

• Advantages:• Very easy and fast to set up

• Disadvantages:• No advanced publication features

SEO/WEB METADATA

• By default SharePoint pages include little metadata (search engines)

• Publishing + 2013 allows us to provide more data• Description

• Keywords

• Sitemap

• Author

• Browser title

• …

FACETS

• Improve search

• Located on the left with search results

• Use metadata (all)

• You can (must!) customize it

• Very efficient to filter out not relevant data and drill down to the right result

DOCUMENT CENTER

•Organizes content with taxonomy• Drop off library

• Applies governance rules based on taxonomy

SITE METADATA NAVIGATION

•Build top navigation + urls with taxonomy

•Useful for SEO

•Great to build a portal in a “semantic” way•Structure depends on your thesaurus

OTHER METADATA

• Author

• Security

• Views (count)

• Containers (sites, doc libs…)

• Resource type (extension, mime)

•Dates (creation, modification…)

• Status (approval)

OK, WE GOT ITBut what is this all about?

VISIONThat’s the part where we have a plan

METADATA LIFECYCLE

Search

Thesaurus

DataNavigation

Governance

Metrics

Tagging/ClassificationStructures

Definition

Improvement

WHAT’S THE POINT?

The general idea is to have a precise representation and a good structure of the ideas of

the company

Copyright Vincent BIRET ;-)

YES BUT!

Users never fill up metadata!

Solutions:

Index their salary on metadata fill up rate

Wall of shame of worst metadata “filler” of the company

Automate that part!

GRAPH? DELVE?Only buzz words?

GRAPH

• First of all it’s a mathematics model• Nodes, edges and by transivity “routes”

• Then data model• Idea that on piece of data is related to another

• Finally Microsoft is making huge investments for businesses• Actors, signals, objects

• Schema is extensible to add METDATA!

DELVE

It’s only one of the clients

DEMODelve portal

WHAT’S THE IDEA?

• Pro-active content

• Discoverability

• No need to search for something anymore

• If we still need to search something, way more efficient

• Make you save time, therefor money

CONCLUSIONTime to say good bye :’-(

CONCLUSION

•We improved navigation

•We improved findability

•We improved discoverability

•We improved user experience

•Everybody saves time

•The company saves money

THANK YOUEVENT SPONSORS

We appreciated you supporting the New York SharePoint Community!

• Diamond, Platinum, Gold, & Silver have tables scattered throughout

• Please visit them and inquire about their products & services

• To be eligible for prizes make sure to get your bingo card stamped by ALL sponsor

EVALUATIONS

QUESTIONS?/ THANKS!