Stream

Join this community to post or comment

Aaron Bradley
owner

Discussion  - 
 
"Hyperlinks are dumb."

"Literally dumb. They link one document to another – one little piece of text to another, but they don’t know how. They don’t know why. They don’t know diddly squat. The only real sense that search engines can make of their incoherent connections is that one page is more popular than another, probably, maybe."

A fabulous short piece from +Cody Burleson that clearly explains the difference between everyday hyperlinking and linked data.

Via +Kingsley Idehen via +Teodora Petkova on Google Semantic Search©®™, and well worth re-sharing here.
Hyperlinks are dumb. Literally dumb. They link one document to another – one little piece of text to another, but they don’t know how. They don’t know why. They don’t know diddly squat. The only real sense that search engines can make of their incoherent connections is that one page is more popular than another, […]
18
9
Barbara Starr's profile photoSEO's profile photo
Add a comment...

Barbara Starr

Discussion  - 
 
Define: ?x and What is ?x yielding different results in organic search vs answer box?

I have not seen this before.  And yes, I was signed in.  

#answersnippettext       #whatisquestiontype   #google #define :
7
2
Barbara Starr's profile photoHart GA's profile photo
Add a comment...

Aaron Bradley
owner

Google Knowledge Graph and Knowledge Vault  - 
 
A featured snippet accordion

Over on Twitter +Dan Leveille posted a screenshot of Google query results for "pet crab" (http://bit.ly/1LHAbwU), and this list of featured snippets was independently noted by +Ajay Prasad (http://bit.ly/1LHAzvf).

I'm calling this a "list of featured snippets" because that's what it appears to be, with some nuances.

- The questions under "People also ask" are indeed questions of the sort that typically return featured snippets, rather than related queries (those related searches listed at the bottom of the page are quite different than what's found on this list).

- The featured snippets in the accordion include a link to search for the question that forms the snippet title.  This makes sense as, unlike a stand-alone featured snippet (e.g. http://bit.ly/1LHBc8c), the accordion snippets we're triggered directly by the user's query.

At the end of the day I see this fundamentally as a user interface change more than anything else, but there's definitely insights to be derived from the accordion list of questions because AFAIK this is the first time we've seen Google generate a list of related questions rather than a list of related searches which, as per the first point above, are quite different things.

Dan in his tweet frames this as "Google is now recommending related #KnowledgeVault answers", but I'd be hard-pressed to say that this derives from the Knowledge Vault which, you'll recall, is "a Web-scale probabilistic knowledge base that combines extractions from Web content (obtained via analysis of text, tabular data, page structure, and human annotations) with prior knowledge derived from existing knowledge repositories."  (http://bit.ly/1tDaS4Z)  That is, a fact engine  (still AFAIK unverified in the wild) that would be more likely to return Knowledge Graph-type results in the form of direct factual answers and/or a Knowledge Panel (e.g. http://bit.ly/1LHCkbZ).

As per the recent SEL article on featured snippets from +Eric Enge (http://bit.ly/1LHCkbZ) the key here in being featured in one is "good, user-centered page design."

A note that I showed the collapsed and expanded forms of the accordion to note:

- The variety of sources drawn upon, including both commercial websites and Wikipedia
- The presence of petsmart.com in both the featured snippets and the SERPs (an otherwise sometimes-seen feature of featured snippets reinforced here.

Finally, as others have noted .... "pet crab"? :)  FWIW I haven't been able to generate this for any other "pet [insert animal name here]" query, which makes the oddness of this result even odder IMO (the example query from Ajay is far more pedestrian).

#featuredsnippets   #google
14
11
Alex C.'s profile photoNasah Zined's profile photoBarbara Starr's profile photoHart GA's profile photo
5 comments
 
+Aaron Bradley Maybe I am missing something tho? as "featured snippets" are perhaps different?  I was thinking perhaps these may be 2 differently maintained indexes at first, but definitely need to process this further before making any definitive comment
Add a comment...

Jan-Willem Bobbink

Google Knowledge Graph and Knowledge Vault  - 
 
Learning to Explain Entity Relationships in Knowledge Graphs
3
Aaron Bradley's profile photo
 
Great share, thanks J-W!
Add a comment...

Aaron Bradley
owner

Discussion  - 
 
"Product data submitted to Manufacturer Center influences how your brand is represented across Google"

Among the many interesting things about this initiative is the breadth of its impact:  not just Google shopping, but other Google properties as well - including, apparently, Google search.  Emphasis mine:

"Manufacturer Center allows manufacturers to provide better and more accurate information about their products to improve the user experience on Google.com, Google Shopping, and other Google services." (http://bit.ly/1RQeAaL)

At the heart of Manufacturer Center is product data (that can provided in a number of formats, and modified by a number of mechanisms) consisting of required and optional product attributes.  There's much here that's similar to other feed formats, to the properties supported by schema.org/Product, and the nascent GS1 Web Vocabulary.

Half the required fields are all about precise product identification and disambiguation (http://bit.ly/1RQg3xF):  ID, unique product identifier (GTIN) and manufacturer part number (MPN).

All in all, in seems reasonable to assess that Google's primary purpose here is to disambiguate products in order to more easily compare apples to apples, and provide consumers with purchase and vendor options rather than spreading information about a single product over several listings.

Structured snippets writ large.

Remember those big Knowledge Panel structured-snippet type PLAs with price comparisons that have re-emerged lately, as reported by +Jennifer Slegg (http://bit.ly/1J2CYff)?  Expect more of these.

There's also the possibility that we could start to see such types of snippets in organic search.  Well, eventually I don't see how we can expect anything else.  Why show 10 blue links when all the attributes of the actual product are the same?

#structureddata   #google   #gtin   #products  
10
3
Michael Andrews's profile photoAaron Bradley's profile photoEmanuele Vaccari's profile photoJomer Gregorio's profile photo
6 comments
 
Find this interesting for a couple of reasons.  Seems to be a proprietary Google metadata scheme: wonder if others might use it for other purposes. It's not an alternative to schema.org if one wants to have coverage on other search engines.   And interesting that the information is not published through HTML markup or a JSON script, but using RSS.
Add a comment...

Aaron Bradley
owner

Discussion  - 
 
"Interested in learning how to take taxonomies to the next level?"

Developing your own taxonomies and ontologies can open up numerous possibilities for leveraging semantic technologies for marketing purposes (especially content marketing).

In this webinar, for example, the presenter will "show us how to take it further and extend the ontology to include how consumers search on the web and how they can be targeted on Facebook for advertising purposes."

I thought this would be of particular to Community members, as so many of the SEOs here have a local focus.

#taxonomy   #ontology   #localbusiness  
Click here to register (Registration closes 5 PM ET, August 11, 2015). Date: August 13, 2015. Time: 1:00pm ET. DESCRIPTION. Interested in learning how to take taxonomies to the next level? Join presenter Michael Bauer as he shows us how to take a small and medium-sized enterprises (SMEs) ...
5
Jarno van Driel's profile photo
 
Registered. Thanks. 
Add a comment...

Barbara Starr

Discussion  - 
 
Buy Now - Near me

Looking at Google’s announcement today: http://adwords.blogspot.com/2015/07/winning-shopping-micro-moments.html

I have been anticipating the formal/official introduction/announcement of the  Google “buy now” button and as to how they may initially introduce it.  I was both surprised and intrigued by the fact that they chose to do so via a mechanism that was perhaps predictable via prior info, however the fact that it emphasized “local search” (to use seo terminology here) seriously surprised me.  On the other hand, given their monetization model (primarily  ad based vs listing based) and the  lack of the fact that CSEs (and their frustration) in terms of their current ability to tap
 the “local” market makes it a brilliant starting point (reckon it is the tip of the iceberg.

Brilliant move on Google's part (my POV)


I have alluded to its imminent introduction and potential mechanism here in the past:


Earlier relevant posts:

https://plus.google.com/+BarbaraStarr/posts/cQxzV3sgcj9

https://plus.google.com/u/0/+BarbaraStarr/posts/LrtbF5fBUqp

And also in this presentation at a recent san diego semantic web/seo meetup: 

https://plus.google.com/+MarkKellerConsulting/posts/YWVRxEZ8546

which I gave with  +Bill Slawski 

#buynowbutton   #google   #localsearch   #google  
9
8
Borislav Arapchev's profile photogeorge oloo's profile photoTyler Harding's profile photoBarbara Starr's profile photo
2 comments
 
Most importantly, IMHO, "the service will also be able to access a user's browsing history to show if the price of a product a shopper has searched for in the past has fallen."
Add a comment...

Aaron Bradley
owner

Tools  - 
 
https://en.wikipedia.org/wiki/XML -> http://dbpedia.org/page/XML

Of late the very handy Wikipedia to DBpedia bookmarklet created by +Bob DuCharme hasn't been functioning, due to Wikipedia's shift from HTTP to HTTPS.  This updated bookmarklet simply restores the functionality of the original.

#wikipedia   #dbpedia   #structureddata   #bookmarkleets  
Bookmarklet to display the DBpedia equivalent of an English Wikipedia page.
1
Bob DuCharme's profile photoAaron Bradley's profile photo
2 comments
 
Thanks +Bob DuCharme!
Add a comment...

Aaron Bradley
owner

schema.org  - 
 
How to create a recipe that isn't using schema.org

When +John Simonson asked "What if the recipe is for something else, such as a fly tying pattern recipe?  What schema type do you use?" the answer from +Thad Guidry - "'Things that have steps' can use http://schema.org/ListItem" - got me thinking about schema.org/Recipe could be used for other types of "recipes".

This in turn led to me to put together "How to Make a Rock Garden" (perhaps my best work of '15 so far) to see how it played out:  the JSON-LD is below.

Curiously enough, as I was wrangling the code +Vicki Tardif Holland responded to Thad's suggestion (see callout link), noting that this had come up before and:

"It would be nice to have a general Instructions or HowTo type, with Recipe as a subtype, but the current usage of Recipe and its properties strongly imply instructions for cooking."

Just having gone through the "How to Make a Rock Garden" exercise I can attest to the fact that Recipe is basically ill-suited to such purposes.  And leveraging structured data to help generate the step-by-step instructions that have now long been appearing in the SERPs (e.g. http://bit.ly/1LiVbtv) has long been a topic of discussion in the search marketing world.

So a hefty +1 for a HowTo or Instructions schema - which I confidently predict would excellent adoption right off the bat.

<script type="application/ld+json">
{
  "@context":"http://schema.org",
  "@type":"Recipe",
  "name": "How to Make a Rock Garden",
  "description": "Rock gardens are great because they don't require watering.",
  "recipeIngredient": [
    "1 to 100 rocks",
    "A bunch of sand"
  ],
  "prepTime": "PT1H",
  "recipeInstructions": {
    "@type":"ItemList",
    "numberOfItems": 2,
    "itemListElement": [
      {
        "@type":"ListItem",
        "position": 1,
        "item": "Rocks"
      },
      {
        "@type":"ListItem",
        "position": 2,
        "item": "Sand"
         }
       ]
     }
   }
  "recipeYield": "1 rock garden"
}
</script>
From : Vicki Tardif Holland Date : Wed, 15 Jul 2015 11:44:23 -0400. Message-ID : To : Thad Guidry Cc : john@webstreamdynamics.com, "schema.org Mailing List" ...
1
Carrie Hill's profile photo
Add a comment...
 
Isn't JSON-LD hidden markup?

I'm a bit confused with Google's structured markup rules at the moment. Their guidelines state you should not use hidden markup except in a few special cases.

https://developers.google.com/structured-data/policies

But JSON-LD is completely hidden!

What gives?
4
Martin Hepp's profile photoAaron Bradley's profile photo
21 comments
 
+Martin Hepp Your moment of celebration well-earned!  I'm very happy the writing on the wall indeed turned out to be ... the writing on the wall. :)
Add a comment...

About this community

A community for the discussion of topics at the intersection of search engine marketing and semantic web technologies This includes - but is not limited to - schema.org and means of marking up HTML with structured data, such as microdata, RDFa and JSON-LD. While "search marketing" is part of this community's name, discussions about digital marketing and the semantic web outside of the context of SEO and PPC are also welcome and encouraged. New to this Community? Select "Introductions" as the post type and tell us something about yourself! Want to share code examples? We like jsbin.com (but use what you like). Share the word! Short link for this Community: http://bit.ly/semsearch #schema #schemaorg #rdfa #microdata #knowledgegraph #jsonld

Aaron Bradley
owner

schema.org  - 
 
"There should be a type for describing instructions outside of recipes"

SERP feature watchers will be very interested in this "how to" schema.org type when and if it rolls out, as it'll obviously at least support the provision of "how to" information directly to the search engines, whereas the basis on which Google et al. currently identifies such step-by-step instructions and returns them directly in the SERPs is fairly opaque.

If you want to pitch in...

<div itemscope itemtype="http://schema.org/HowTo">
  <ul itemprop="step" itemscope itemtype="http://schema.org/ListItem">
  <li itemprop="item">Go to https://github.com/</li>
...
 :)
This has come up a few times, most recently on public-schemaorg@w3.org. There should be a type for describing instructions outside of recipes. There are a lot of how-to videos and websites that wou...
8
1
Wolfgang Anton Jagsch's profile photoMichael Andrews's profile photoJomer Gregorio's profile photo
6 comments
 
I'd echo the comment in Git about looking at DITA, since it is widely used in tech documentation/online help already and I advocate having external markup being aligned with internal content structures as much as possible.  The two different types of instructions are set-up and trouble shooting.  

There are also a new experience API working group looking at learning-oriented instructions, which looks like it will be developing JSON-LD vocabularies https://www.w3.org/community/xapivocabulary/ .  
Add a comment...

Jeewan Garg

Discussion  - 
 
Google Business Page Review on Website

Can I use Google business Page review for my website. And use schema to show in Google search result.

If yes then can somebody guide me how to do it. 
3
Jarno van Driel's profile photoJeewan Garg's profile photo
2 comments
 
Thank you +Jarno van Driel for your guidance. I have installed the code. May I request you to please check it up & tell me if i miss any thing.. my website is : jeewangarg.com 
Add a comment...

Site Raiser

Introductions  - 
 
Hello
1
Site Raiser's profile photo
7 comments
 
This is the schema.org breadcrumb generated for that code example, lol!
<nav><div style="color:#aaa;" itemscope="itemscope" itemtype="http://data-vocabulary.org/Breadcrumb">
<span class="breadcrumb-lead">YOU ARE HERE: </span><a class="internal" title="Go to Homepage" href="/" itemprop="url"><span itemprop="title">Homepage </span></a>> <div itemprop="child" itemscope="itemscope" itemtype="http://data-vocabulary.org/Breadcrumb" style="display: inline;"><a class="internal" title="Go to Products-and-Services" href="/products-and-services" itemprop="url"><span itemprop="title">Products-and-Services</span></a> > <div itemprop="child" itemscope="itemscope" itemtype="http://data-vocabulary.org/Breadcrumb" style="display: inline;"><a class="internal" title="Go to Web-Development" href="/products-and-services/web-development" itemprop="url"><span itemprop="title">Web-Development</span></a> > <div itemprop="child" itemscope="itemscope" itemtype="http://data-vocabulary.org/Breadcrumb" style="display: inline;"><a class="internal" title="Go to Code-Examples" href="/products-and-services/web-development/code-examples" itemprop="url"><span itemprop="title">Code-Examples</span></a> > <div itemprop="child" itemscope="itemscope" itemtype="http://data-vocabulary.org/Breadcrumb" style="display: inline;"><a class="internal" title="Go to Breadcrumb-Generator" href="/products-and-services/web-development/code-examples/breadcrumb-generator" itemprop="url"><span itemprop="title">Breadcrumb-Generator</span></a></div></div></div></div></div></nav>
Add a comment...

Aaron Bradley
owner

Rich and structured snippets  - 
 
Yes Virginia, Google rich snippets can be generated from JSON-LD

I've little time to say much about this right now aside from "hooray!" and thanks to +Grace Massa Langlois for her efforts that have resulted in this finding.

#jsonld   #richsnippets   #google   #schemaorg  
Recipe rich snippets in Google based on data provided via JSON-LD have now been confirmed in the wild, thanks to Grace Massa Langlois.
24
5
Site Raiser's profile photoMarine Di Paolo's profile photoOleg Baskov's profile photoFlorence Auréart's profile photo
10 comments
 
Thanks +Dan Brickley (I had noticed that yesterday and updated the post accordingly).  And if that was in response to this discussion and/or the post that came out it, congratulations on doubtlessly setting some sort of speed record for getting Google documentation updated. :)
Add a comment...

Martin Hepp

Rich and structured snippets  - 
 
FYI: I just described a proper way for modeling product variants with multiple prices in schema.org
6
1
Jarno van Driel's profile photoMartin Hepp's profile photoMarine Di Paolo's profile photo
7 comments
 
I actually think a subtype of ItemList would be better But the reason for my recipe is that it 1. is conceptually sound (other than the WebPage -> offers -> Offer -> itemOffered pattern, which abuses the offers property), yet still works with the existing Google handling of price ranges for multiple offers.

We could add a new type to schema.org, but we would then have to hope and wait for Google to accept this in the very same fashion. So I favor to simply use my recipe at large. It will have the same effect.

In theory, you could also make the main product and detailed product each a schema:Product Model and link them via schema:isVariant.

Note that my pattern is conceptually sound only for variants of the same product, not completely different products. For the latter, there is no supported Google Rich Snippet type anyway.
Add a comment...

Aaron Bradley
owner

Discussion  - 
 
"... to help bring context to a complicated topic"

"Today, The Post began testing a new feature called Knowledge Map.... Knowledge Map gives readers an easier way to catch up on ongoing stories by quickly and seamlessly providing relevant background, additional information or answers to frequently asked questions, when the reader wants it."

I'm willing to be that the panel that comes up when a user clicks on a Map link - very reminiscent of a Google Knowledge Panel or Bing Snapshot - is hard-coded in this instance, but their vision is grander than this.

Engineering Director for Data Science at the Post, Dr. Sam Ham, says:

"This iteration sets us up to use data mining techniques to identify and surface contextual content for our readers. We are also working on parallel applications to drive engagement with our native advertising content. Our ultimate goal is to mine big data to surface highly personalized and contextual data for both journalistic and native content."

#knowledgegraphs   #news   #nlp  
Knowledge Map gives readers an easier way to catch up on ongoing stories by quickly and seamlessly providing relevant background, additional information or answers to frequently asked questions, when the reader wants it.
8
1
Wolfgang Anton Jagsch's profile photoJomer Gregorio's profile photo
 
great share - thanks for that.
Add a comment...

Aaron Bradley
owner

schema.org  - 
 
Coming soon.
5
Add a comment...

Aaron Bradley
owner

Discussion  - 
 
In which +Rand Fishkin talks about machine learning

A lot.  And in a coherent and useful way.  _Ay carumba._

One of the best presentations on "onsite SEO" I've seen in ages, and definitely worth a read.

#machinelearning   #cognitivecomputing   #google   #entities  
8
6
Vincent Velema's profile photoHeather Koroll's profile photo
Add a comment...

Aaron Bradley
owner

Bing Snapshot and Satori  - 
 
"If you really want to hear about it..."

There, you now have only 73,396 words to go before you finish The Catcher in the Rye - according to Bing, and via the tireless +Jennifer Slegg.

I've been searching, without success, for a canonical source for these details.  Anyone know?

By the way, only for US-facing results right now (I had to go to a US proxy to see these results), at least for me (with that acknowledging that Ms. Slegg is also a British Columbian).

#bingsnapshot   #knowledgegraphs   #books
Bing has added some new detail to their book knowledge panels in the search results detailing both word count and reading level. The one drawback to the reading level is that Bing offers no link or reference for searchers to know what AR points or ATOS reading level means.  Unless you are an educator, you …
2
Jennifer Slegg (Jenstar)'s profile photo
2 comments
 
I also added the source to for the word count and reading level to the article.  It is AR Bookfinder.
Add a comment...
 
Latest update on +WordLift v 3.0 - Creating a Linked Open Dataset using +WordPress. The talk was organised by the +Università La Sapienza in the context of the Lider Project (http://www.lider-project.eu/ - Linguistic Linked Data cloud). 
8
2
Aaron Bradley's profile photoAndrea Volpini's profile photoDavid Riccitelli's profile photoTeodora Petkova's profile photo
2 comments
 
Thanks +Aaron Bradley ping me and I'll give you a preview of this latest release (that is not yet out in the wild) anytime.  
Add a comment...