Exploiting microformats for SEO purposes

1

Exploiting microformats and

semantic WEB for SEO purposes

SALVATORE CAPOLUPO

2

ABOUT ME

Computer Engineer PHP/MySQL Freelance

Operational Research PhD Tech Blogger

Affiliate-marketer Hosting reseller

3

BACKGROUND

• SEO basics• Websites basics

– i.e. HTML/PHP• Web-server

configuration – i.e. .htaccess

• Keyphrase = meaniningful sequence of keywords

• Word of mouth = the passing of information from person to person

4

4 SIMPLE IDEAS(1) Exploiting Semantic URLs(2) Long-tail Keyphrases(3) Expose product features (e-commerce)(4) Give users chance to judge your product

5

EVOLUTION

http://www.w3.org/People/Ivan/CorePresentations/Applications/


6

STRUCTURE RECOGNITION

7

EXAMPLE

→

→ Which one is more representative? → Google tries in general to “guess”

8

QUALITY OF A WEB-SITEhttp://googlewebmastercentral.blogspot.com/

“More guidance on building high-quality sites”Friday, May 06, 2011 at 11:22 AM

http://googlewebmastercentral.blogspot.com/

9

SCENARIO

SITE-TRUST Would you trust the information presented in this article?

10

#1 : SEMANTIC WEB FOR WEBMASTERS

Sem WEB is a “web of data”

EXAMPLE: Dublin Core metadata record can describe physical resources such as books, digital materials such as video, sound, image, or text files.

11

#2 : SEMANTIC WEB & SEARCH ENGINES

1) Related content into SERPS2) No exact keyword matching!

12

SEMANTIC WEB & SEARCH ENGINES

Google tries to “guess” relevancy of documents using some semantic features

13

LATENT INDEX SEMANTIC

From exact KW matchings to semantic domains

Semantic Domain

What this page is really about ?

Find related terms to a main keyword

14

FACING BLACK BOX

Semantics → give real sense to your contentsExpose relevant data (i.e. price and essential

features of products you sell)

15

APPROACHES

Semantics → spread keyword on page using contextual associations (i.e. cheap car is often used car)

Expose relevant data → integrate HTML with microformats

16

“SEMANTIC” BACKLINKS

SEO = just “inflate” your PR → reducted impact, narrow vision

Automated backlinks (little blogs, random forums, paid links, …)

Semantic-backlinks example:

Comment a site / Be cited by a site pertinent and contextual

FAST & FAST & BADBAD

SLOW & SLOW & GOODGOOD

17

TIPICAL PROBLEMS• Spam• Similar Articles• Content Farms• Duplicate Contents...

18

TIPICAL PROBLEMS• ...Well-knows competitors take the lion's share• → Many researches with:

– Bad Precision – Bad Recall

19

Query entity detection

Search engine receives a search query, determines whether the received search query includes an entity name, and determines whether the entity name is associated with a common word or phrase. Generate a link to a rewritten query or include a restrict identifier associated with the entity name (US Patent 7,536,382)

http://www.makebeliefscomix.com/Comix/

GOOGLE PATENT

20

DON'T REPEAThttp://mdrseo.com

– “echo back” effect - i.e. similar low-quality contents about same topic:

I evaluate X about Y

We think X about Y

I (!) suppose X about Y

WE ARE SURE: X=Y

21

SEO

• Documents• Links

22

SEO → SEO++

• Documents → data, things, entities• Links → “relationships”

page

entity

sem-rel

23

THE PRICE TO PAY

• Extend keyword using semantic domains

– Microformats → little-impact HTML changes

24

WEB 3.0• RDF = “general method for

conceptual description of information that is implemented in web resources”

• µF = re-use existing HTML/XHTML tags to convey meta-data and other attributes in web pages and/or RSS

25

REMARKS

– QUALITY ≈ TRUST concept– Latent Index Semantic– Relevant details to search engines:

• Microformats– Avoid “echo back” effects

26

my-site.com/index.php?sect=4&p=2

SEMANTIC URL = “human-readable”

27


e-commerce.com/buy-php-mysql-hosting buythis.net/hosting/cheap


28

SEMANTIC URL = “human-readable” easily remembered; navigational structure; intuitively manipulated; more durable; more bookmarked; may improve security!

29

REMARKS GOOGLE DOESN'T USE EXACT MATCHING

Example: buyshoes.com/cheap-buy/low-cost-shoes KEYWORD STUFFING

IS NOT USEFUL FOR USERS → NOT USEFUL FOR SEARCH ENGINES

30

titledescriptionURL

URLs ARE NOT ALL...

“When searchers scan through search engine listings what usually they read?”

43%: descriptions 30%: titles 21%: URLs

http://yourseoplan.com

31

OPTIMIZING TITLE/DESCRIPTION

Semantically related means:

Taking into account THE MOST STRICTLY LINKED UP terms

32

INTRODUCING: LONG TAIL

Searchers often use too generic stems (i.e. hostel, site-builder)

In others cases people tends to be more specific (i.e. london cheap hostels, free site-builder mac)

33

THE LONG TAILLong tail addresses usually many specifical keywords (niches)

“The future [...] is in the millions of niche markets at the shallow end of the bitstream”

Image by: www.online-marketing.it

http://www.wired.com/wired/archive/12.10/tail_pr.html

CHRIS ANDERSEN

http://www.online-marketing.it/


34 http://www.seomoz.org/blog/illustrating-the-long-tail

http://www.seomoz.org/blog/illustrating-the-long-tail

35

EXAMPLES: GOOGLE WHEEL & ADWORDS KW TOOL

36


37

SEO, SEO service(s), SEO company, company SEO, ...

38

Stupidly simple SEO, black-hat SEO, free SEO tips

39

PROBLEMS

• Usually de-structured informations

40

MICROFORMATS FOR IN-PAGE OPTIMIZATION

Provide well-structured informations– User's reviews– Events– ...

41

Microformats

• Personal Informations (hCard)• Social (FOAF)• Events (hCalendar)• Products (RDFa/GoodRelations)• Reviews (hReview)

42

GoodRelations is a standardized vocabulary for product, price, and company data that:

(1) can be embedded into existing [...] Web pages and (2) can be processed by other computers.

2spaghi.it bestbuy.com comparatorehosting.com

oreilly.com overstock.com ...

http://www.ebusiness-unibw.org/wiki/GoodRelations#Applications

43

GoodRelations + Google

44

GoodRelations Core Classes

• gr:BusinessEntity for a company or business• gr:Offering for an offer to sell, repair, lease something, or to

express interest in such an offer• gr:ProductOrServiceModel for the datasheet describing the

features of a product• gr:LocationOfSalesOrServiceProvisioning for a store or

location from which the offer is available

45


• gr:BusinessEntity for a company or business• gr:Offering for an offer to sell, repair, lease

something, or to express interest in such an offer• gr:ProductOrServiceModel for the datasheet describing the



46

gr:Offering → products tassonomyhttp://www.ebusiness-unibw.org/tools/grsnippetgen/

47

gr:Offering complete example

Descriptive text: We sell (gr:hasBusinessFunction) “Hosting product 123” which “Supports PHP 5, MySQL, Perl, unlimited emails and subdomains”, and is specifical for (gr:eligibleCostumerTypes) Enduser, Business and PublicIstitution. It costs (gr:hasCurrencyValue) 50 $ (gr:hasCurrency), and we accept Paypal, VISA and Mastercard Payments (gr:acceptedPaymentsMethods)

http://www.ebusiness-unibw.org/tools/grsnippetgen/

48

gr:Offering multi-products

If you have multiple products within the same web page, simply create and copy the snippet multiple times

replace all ocurrences of the element #offer inside the same snippet by #offer1, #offer2, etc.


49




features of a product ...• gr:LocationOfSalesOrServiceProvisioning for a store or

location from which the offer is available ...

• See http://www.heppnetz.de/ontologies/goodrelations/v1

http://www.heppnetz.de/ontologies/goodrelations/v1

50

Company: Semantic Representation

← Dictionary specification

51

WHY SHOULD GR BE USEFUL FOR SEO?

• Universal Product Catalogue

• SERP quality improvement

• Well-known format

52

hReview is a microformat for publishing reviews of books, music, films, restaurants, businesses, holidays, etc.[1] using (X)HTML on web pages

It uses HTML classes and rel attributes

shockingexperts.com elevatelocal.co.uk assistenza-clienti.it

hReview

http://wordpress.org/extend/plugins/wp-customer-reviews/

53

hReview + Google

• Properties– fn (formatted name)– Reviewer (GameSpot)– Date (Nov 3, 2009)– summary– description– rating ( )

• Individual Review• Review-aggregate

54

hReview into HTML markup

By default, Google assumes that your site uses a 5-point scale, where 5 is the best possible rating and 1

is the worst.

55

Review-aggregate: example

http://www.google.com/webmasters/tools/richsnippets


56

Another example: Google Recipes

57

WHY SHOULD hReview BE USEFUL FOR SEO?

• Increase word-of-mouth (visits)



58

EQUIVALENTFORMS

RDFa

microformats

microdata

59

THE LAND OF THE UNICORN...?

If everyone would subscribe to such a system and create good metadata for the purposes of describing their goods, services and information →

• it would be a trivial matter to search the Internet for highly qualified, context-sensitive results

60

CRITICISM: “metacrap”

http://www.well.com/~doctorow/metacrap.htm

• There's more than one way to describe something• Dishonesty: emphasize my high-scoring axes && de-

emphasize my low-scoring axes

“A leopard can't change his spots”


61

FAQ

• Microformats “ignored” by Google?– Not Visible to users (i.e. display:none)– Wrong Sintax– Low Relevance– Crawler not yet updated – ...patience ;)

Mmm... no microformats shown in Google...???

62

4-STEP SEO++

• Semantic KW research• Semantic-URLs• GoodRelations for improving your e-

commerce site• Give user the chance to review/vote your

contents (software, services, … )

63

LOOKING AT THE FUTURE

• Naturally adapting semantics in Twitter (µBlogs)• LinkedData: "recommended best practice for

exposing, sharing, and connecting pieces of data, information, and knowledge on the Semantic Web”

• SPARQLE http://en.wikipedia.org/wiki/SPARQL

http://en.wikipedia.org/wiki/SPARQL

64

REFERENCES

http://www.heppnetz.de/projects/goodrelations/http://www.google.com/webmasters/tools/richsnippetshttp://linkeddata.org/

http://www.mattgroening.com/ http://www.makebeliefscomix.com/Comix/

http://www.heppnetz.de/projects/goodrelations/


http://linkeddata.org/

http://www.mattgroening.com/


65

THANK YOU!

http://twitter.com/salcapolupohttp://mdrseo.comhttp://seobynight.blogspot.com

http://asiunical.org← complete transcription

http://twitter.com/salcapolupo

http://mdrseo.com/

http://seobynight.blogspot.com/

http://asiunical.org/

Exploiting microformats and semantic-WEB for SEO purposes2

ABOUT ME

Computer Engineer PHP/MySQL Freelance

Operational Research PhD Tech Blogger

Affiliate-marketer Hosting reseller


BACKGROUND

• SEO basics• Websites basics

– i.e. HTML/PHP• Web-server

configuration – i.e. .htaccess

• Keyphrase = meaniningful sequence of keywords

• Word of mouth = the passing of information from person to person


4 SIMPLE IDEAS(1) Exploiting Semantic URLs(2) Long-tail Keyphrases(3) Expose product features (e-commerce)(4) Give users chance to judge your product


EVOLUTION



STRUCTURE RECOGNITION


EXAMPLE

→

→ Which one is more representative? → Google tries in general to “guess”


QUALITY OF A WEB-SITEhttp://googlewebmastercentral.blogspot.com/

“More guidance on building high-quality sites”Friday, May 06, 2011 at 11:22 AM


SCENARIO

SITE-TRUST Would you trust the information presented in this article?


#1 : SEMANTIC WEB FOR WEBMASTERS

Sem WEB is a “web of data”

EXAMPLE: Dublin Core metadata record can describe physical resources such as books, digital materials such as video, sound, image, or text files.


#2 : SEMANTIC WEB & SEARCH ENGINES

1) Related content into SERPS2) No exact keyword matching!


SEMANTIC WEB & SEARCH ENGINES

Google tries to “guess” relevancy of documents using some semantic features


LATENT INDEX SEMANTIC

From exact KW matchings to semantic domains

Semantic Domain

What this page is really about ?

Find related terms to a main keyword


FACING BLACK BOX

Semantics → give real sense to your contentsExpose relevant data (i.e. price and essential

features of products you sell)


APPROACHES

Semantics → spread keyword on page using contextual associations (i.e. cheap car is often used car)

Expose relevant data → integrate HTML with microformats


“SEMANTIC” BACKLINKS

SEO = just “inflate” your PR → reducted impact, narrow vision

Automated backlinks (little blogs, random forums, paid links, …)

Semantic-backlinks example:

Comment a site / Be cited by a site pertinent and contextual

FAST & FAST & BADBAD

SLOW & SLOW & GOODGOOD


TIPICAL PROBLEMS• Spam• Similar Articles• Content Farms• Duplicate Contents...


TIPICAL PROBLEMS• ...Well-knows competitors take the lion's share• → Many researches with:

– Bad Precision – Bad Recall


Query entity detection

Search engine receives a search query, determines whether the received search query includes an entity name, and determines whether the entity name is associated with a common word or phrase. Generate a link to a rewritten query or include a restrict identifier associated with the entity name (US Patent 7,536,382)


GOOGLE PATENT


DON'T REPEAThttp://mdrseo.com

– “echo back” effect - i.e. similar low-quality contents about same topic:

I evaluate X about Y

We think X about Y

I (!) suppose X about Y

WE ARE SURE: X=Y


SEO

• Documents• Links


SEO → SEO++

• Documents → data, things, entities• Links → “relationships”

page

entity

sem-rel


THE PRICE TO PAY

• Extend keyword using semantic domains

– Microformats → little-impact HTML changes


WEB 3.0• RDF = “general method for

conceptual description of information that is implemented in web resources”

• µF = re-use existing HTML/XHTML tags to convey meta-data and other attributes in web pages and/or RSS


REMARKS

– QUALITY ≈ TRUST concept– Latent Index Semantic– Relevant details to search engines:

• Microformats– Avoid “echo back” effects






e-commerce.com/buy-php-mysql-hosting buythis.net/hosting/cheap



SEMANTIC URL = “human-readable” easily remembered; navigational structure; intuitively manipulated; more durable; more bookmarked; may improve security!


REMARKS GOOGLE DOESN'T USE EXACT MATCHING

Example: buyshoes.com/cheap-buy/low-cost-shoes KEYWORD STUFFING

IS NOT USEFUL FOR USERS → NOT USEFUL FOR SEARCH ENGINES


titledescriptionURL

URLs ARE NOT ALL...

“When searchers scan through search engine listings what usually they read?”

43%: descriptions 30%: titles 21%: URLs

http://yourseoplan.com


OPTIMIZING TITLE/DESCRIPTION

Semantically related means:

Taking into account THE MOST STRICTLY LINKED UP terms


INTRODUCING: LONG TAIL

Searchers often use too generic stems (i.e. hostel, site-builder)

In others cases people tends to be more specific (i.e. london cheap hostels, free site-builder mac)


THE LONG TAILLong tail addresses usually many specifical keywords (niches)

“The future [...] is in the millions of niche markets at the shallow end of the bitstream”

Image by: www.online-marketing.it


CHRIS ANDERSEN

Exploiting microformats and semantic-WEB for SEO purposes34 http://www.seomoz.org/blog/illustrating-the-long-tail






SEO, SEO service(s), SEO company, company SEO, ...


Stupidly simple SEO, black-hat SEO, free SEO tips


PROBLEMS

• Usually de-structured informations


MICROFORMATS FOR IN-PAGE OPTIMIZATION

Provide well-structured informations– User's reviews– Events– ...


Microformats

• Personal Informations (hCard)• Social (FOAF)• Events (hCalendar)• Products (RDFa/GoodRelations)• Reviews (hReview)


GoodRelations is a standardized vocabulary for product, price, and company data that:

(1) can be embedded into existing [...] Web pages and (2) can be processed by other computers.

2spaghi.it bestbuy.com comparatorehosting.com

oreilly.com overstock.com ...

http://www.ebusiness-unibw.org/wiki/GoodRelations#Applications


GoodRelations + Google









• gr:BusinessEntity for a company or business• gr:Offering for an offer to sell, repair, lease

something, or to express interest in such an offer• gr:ProductOrServiceModel for the datasheet describing the




gr:Offering → products tassonomyhttp://www.ebusiness-unibw.org/tools/grsnippetgen/


gr:Offering complete example

Descriptive text: We sell (gr:hasBusinessFunction) “Hosting product 123” which “Supports PHP 5, MySQL, Perl, unlimited emails and subdomains”, and is specifical for (gr:eligibleCostumerTypes) Enduser, Business and PublicIstitution. It costs (gr:hasCurrencyValue) 50 $ (gr:hasCurrency), and we accept Paypal, VISA and Mastercard Payments (gr:acceptedPaymentsMethods)



gr:Offering multi-products

If you have multiple products within the same web page, simply create and copy the snippet multiple times

replace all ocurrences of the element #offer inside the same snippet by #offer1, #offer2, etc.






features of a product ...• gr:LocationOfSalesOrServiceProvisioning for a store or

location from which the offer is available ...

• See http://www.heppnetz.de/ontologies/goodrelations/v1


Company: Semantic Representation

← Dictionary specification


WHY SHOULD GR BE USEFUL FOR SEO?

• Universal Product Catalogue




hReview is a microformat for publishing reviews of books, music, films, restaurants, businesses, holidays, etc.[1] using (X)HTML on web pages

It uses HTML classes and rel attributes

shockingexperts.com elevatelocal.co.uk assistenza-clienti.it

hReview

http://wordpress.org/extend/plugins/wp-customer-reviews/


hReview + Google

• Properties– fn (formatted name)– Reviewer (GameSpot)– Date (Nov 3, 2009)– summary– description– rating ( )

• Individual Review• Review-aggregate


hReview into HTML markup

By default, Google assumes that your site uses a 5-point scale, where 5 is the best possible rating and 1

is the worst.


Review-aggregate: example



Another example: Google Recipes


WHY SHOULD hReview BE USEFUL FOR SEO?

• Increase word-of-mouth (visits)




EQUIVALENTFORMS

RDFa

microformats

microdata


THE LAND OF THE UNICORN...?

If everyone would subscribe to such a system and create good metadata for the purposes of describing their goods, services and information →

• it would be a trivial matter to search the Internet for highly qualified, context-sensitive results


CRITICISM: “metacrap”


• There's more than one way to describe something• Dishonesty: emphasize my high-scoring axes && de-

emphasize my low-scoring axes

“A leopard can't change his spots”


FAQ

• Microformats “ignored” by Google?– Not Visible to users (i.e. display:none)– Wrong Sintax– Low Relevance– Crawler not yet updated – ...patience ;)

Mmm... no microformats shown in Google...???


4-STEP SEO++

• Semantic KW research• Semantic-URLs• GoodRelations for improving your e-

commerce site• Give user the chance to review/vote your

contents (software, services, … )


LOOKING AT THE FUTURE

• Naturally adapting semantics in Twitter (µBlogs)• LinkedData: "recommended best practice for

exposing, sharing, and connecting pieces of data, information, and knowledge on the Semantic Web”

• SPARQLE http://en.wikipedia.org/wiki/SPARQL


REFERENCES

http://www.heppnetz.de/projects/goodrelations/http://www.google.com/webmasters/tools/richsnippetshttp://linkeddata.org/

http://www.mattgroening.com/ http://www.makebeliefscomix.com/Comix/


THANK YOU!

http://twitter.com/salcapolupohttp://mdrseo.comhttp://seobynight.blogspot.com

http://asiunical.org← complete transcription

Exploiting microformats for SEO purposes

Technology

pay extend

shoes keyword

yi evaluate

echo back

visa andmastercard

millionsof

approachessemantics

exact kw matchings