Building Entities & Connections in 2020
•
16 likes•3,914 views
Entities are machine classified objects and concepts that allow for reliable results, more efficient fact delivery, better UX of web search, and so much more. Basically put - entities are breathtaking! Izzi's talk touches upon entities and the Knowledge Graph within the context of Search over the past 8 years, and how brands and domains can leverage their power to be present and well-represented across SERPs, the risks of doing so, and what this all means for the ecosystem of search.
Recommended
More Related Content
What's hot
What's hot (20)
Similar to Building Entities & Connections in 2020
Similar to Building Entities & Connections in 2020 (20)
More from Izzi Smith
More from Izzi Smith (10)
Recently uploaded
Recently uploaded (20)
Building Entities & Connections in 2020
- 3. 3 Starting my presentation the normal way Starting my presentation with memes and bad jokes
- 4. ⏰
- 5. 5
- 6. 6
- 9. 9 ● Person, place, or thing ● Abstract concepts ● Brands ● Domain ● Identified by machines as totally individual, e.g. with a ● Unique Resource Identifier (URI) ● Wikipedia ID
- 11. 11
- 12. 12
- 13. 13
- 14. 14
- 15. 15
- 16. 16
- 17. In order to exist, a named entity has several requirements ● Name: An entity can have only one ID but multiple names or aliases (e.g. Barack Obama and President Obama) ● Unique identifier: Such as Wikipedia page IDs or URIs (Unique Resource Identifier) ● Type: A way of grouping entities based on categories, organized in a hierarchical structure, e.g. Author > Person ● Attributes: or characteristics most likely to be seen within the Type. E.g. birth date, height, spouse, for a Person. ● Relationships: The links between entities, and their connections.
- 19. Name - Izzi Smith - Birth name: Isobel Smith Unique ID - URL: https://izzismith.com - Social alias: @izzionfire Type - Person Attributes - Day of Birth: 18.08.1993 - Place of Birth: Burnley, UK - Height: 6ft - Job title: Technical SEO Analyst Relationships - Works for: Ryte GmbH - Speaker at: Women in Tech SEO Festival
- 22. ???
- 26. 26 “Aleyda Solis is a a Global SEO Consultant and founded the agency Orainti.” “The Women in Tech SEO Festival is an event in London.” “Izzi Smith seeks beer.”
- 29. Speculative Fiction Is Author of Is Based on actor Starred in Ottawa, Canada Born in 18.11. 1939 has Birth date Dystopian Fiction May, 2003 Published on L.A., USA
- 30. Speculative Fiction Is Based on actor Starred in Ottawa, Canada Born in 18.11. 1939 has Birth date Dystopian Fiction May, 2003 Published on L.A., USA Is Author of
- 31. Speculative Fiction Is Author of Is Based on actor Starred in Ottawa, Canada Born in 18.11. 1939 has Birth date Dystopian Fiction May, 2003 Published on L.A., USA
- 32. Speculative Fiction Is Author of Is Based on actor Starred in Ottawa, Canada Born in 18.11. 1939 has Birth date Dystopian Fiction May, 2003 Published on L.A., USA https://www.theverge.com/2016/10/4/13122406/google-phone-event-stats
- 33. 33
- 34. 34
- 35. 35
- 36. Speculative Fiction Is Author of Is Based on actor Starred in Ottawa, Canada Born in 18.11. 1939 has Birth date Dystopian Fiction May, 2003 Published on L.A., USA
- 37. Speculative Fiction Is Based on actor Starred in Ottawa, Canada Born in 18.11. 1939 has Birth date Dystopian Fiction May, 2003 Published on L.A., USA Is Author of
- 38. How it works in a nutshell: ● The score is always relative to the contextual document surrounding it. ● Determines what the most important topics in a document are ● Factors impacting an entity’s salience score include: ● Position within text, grammatical role, clarity, reference counts, etc. https://cloud.google.com/natural-language
- 40. 40 Unique ID (kg:/x/xxxxx/) Use for future KG API pulls
- 44. 👀 44
- 46. 46 Search System Index Database Search Engine Indexing Engine Ranking Engine Network Query Results Query Document results Izzi’s Device
- 47. 47
- 49. 49 Search System Index Database Search Engine Indexing Engine Ranking Engine Network Query Results Query Results Izzi’s Device Related Entity Database Related Entities Engine
- 51. 51
- 53. Search System Index Database Search Engine Indexing Engine Ranking Engine Related Entity Database Related Entities Engine “Query” 1. During the indexing process, entities are extracted using NLP and Information Extraction. 2. During the ranking process, queries are assessed to be related to any entities (and if there is the need) and the connected data can be provided. 53
- 54. Search System Index Database Search Engine Indexing Engine Ranking Engine Related Entity Database Related Entities Engine “Entity Related Query” ● An Answer ● Structured Information ● Properties of the Entity ● Related Entities ● Related Queries & Answers ● Varying content formats ● A heck load of other stuff 54
- 57. 57 Wow
- 58. Quick points: Entity based search is much more efficient for Search Engines, mainly due to language agnostic retrieval. MFI arguably helped Google properly re-classify and index existing entities and locate new ones. Google is transitioning from a Link Graph to the entity-fuelled Knowledge Graph Search will (and already does) rely on contextual and device specific signals https://mobilemoxie.com/blog/entity-first-indexing-mobile-first-crawling-1-of-5/
- 60. ツ
- 61. 61
- 62. Entity-Related Searches = 0 Click SERPs ● People don’t want / need to parse your content heavy document ● Google is powerful at providing answers to Entity-Related Searches ● Adapt with the times by optimizing your owned entities! ● Or move on and accept your loss. Which is fine! https://searchengineland.com/now-more-50-of-google-searches-end-without-a-click-to-other-content-study-find s-320574
- 67. 4. And so on... 67
- 69. Featured Snippets with Extracted Entities 69
- 70. Knowledge Cards with supplemental Featured Snippets 70
- 72. 72 Searcher makes informational query Satisfaction is met :)Google directly answers People carry out more actions on Google People trust in their algorithms (your ranking) Your brand awareness from entity results Advances in Search technology for our benefit Faster conversions from on-SERP integrations Decreased trust in paid results?? 😏
- 73. 73 Entities
- 78. 78 🗣 Brand Entities 🏠 Domain Entities ● Build up an authoritative voice in your industry ● Ensure your Knowledge Graph results are well-optimized ● Provide a better factual representation of your brand and attributes ● Be seen as a trusted location of information in the web ● Build up more external connections in the web that help your landing pages rank for associated queries ● Structure data with Schema for better machine classification and interpretation 💡 Topics Entities ● Find new, valuable content opportunities ● Feed factual information that can be useful outside your native language ● Be present in all types of search facets and features ● Enrich device-agnostic search results E.g. Izzi Smith E.g. https://izzismith.com E.g. “SEO memes”
- 80. 80
- 81. 81
- 82. 82
- 83. 83
- 84. 84
- 85. 😏 85
- 86. 😍 86
- 87. 87
- 88. IZZI 88
- 89. Branded Knowledge Result Optimization: ● Check any Entity-Featured Snippets and ask for inclusion ● Claim your Knowledge Panel via GSC ● Even if it’s unlinked / affiliate - it’s still worth it! ● Add Organization/ Business specific Structured Data ● Claim, enrich and monitor your GMB local listings ● Use a square-focus logo that stands out from the pack
- 91. 91
- 92. 92
- 95. 95
- 97. The Process (in a nutshell) ● The system then generates a natural language query to remedy that ● Based on conventions of related entities, the question is raised when there are value mismatches ● Likely predictions are returned and evaluated with query processing. ● Highly confident answers enrich and grow the Knowledge Graph. More info: http://www.seobythesea.com/2018/10/how-googles-knowledge- graph-updates-itself-by-answering-questions/
- 98. 98
- 99. <script type="application/ld+json"> { "@context": "https://schema.org", "@type": "ExhibitionEvent", "name": "Women in Tech SEO Festival", "@id": "https://www.womenintechseo.com/conference", "sameAs":"https://www.facebook.com/groups/698643063902454", "startDate": "2020-03-06", "location": { "@type": "Place", "name": "Barbican Centre", "address": { "@type": "PostalAddress", "addressLocality": "London", "postalCode": "EC2Y8DS" } }, "performer": { "@type": "Person", "name": "Izzi Smith", "worksFor": "Ryte GmbH", "jobTitle": "Technical SEO Analyst" } } } </script> 99
- 100. 100 <script type="application/ld+json"> { "@context": "https://schema.org", "@type": "ExhibitionEvent", "name": "Women in Tech SEO Festival", "@id": "https://www.womenintechseo.com/conference", "sameAs":"https://www.facebook.com/groups/698643063902454", "startDate": "2020-03-06", "location": { "@type": "Place", "name": "Barbican Centre", "address": { "@type": "PostalAddress", "addressLocality": "London", "postalCode": "EC2Y8DS" } }, "performer": { "@type": "Person", "name": "Izzi Smith", "worksFor": "Ryte GmbH", "jobTitle": "Technical SEO Analyst" } } } </script>
- 101. 101 <script type="application/ld+json"> { "@context": "https://schema.org", "@type": "ExhibitionEvent", "name": "Women in Tech SEO Festival", "@id": "https://www.womenintechseo.com/conference", "sameAs":"https://www.facebook.com/groups/698643063902454", "startDate": "2020-03-06", "location": { "@type": "Place", "name": "Barbican Centre", "address": { "@type": "PostalAddress", "addressLocality": "London", "postalCode": "EC2Y8DS" } }, "performer": { "@type": "Person", "name": "Izzi Smith", "worksFor": "Ryte GmbH", "jobTitle": "Technical SEO Analyst" } } } </script>
- 102. 102 <script type="application/ld+json"> { "@context": "https://schema.org", "@type": "ExhibitionEvent", "name": "Women in Tech SEO Festival", "@id": "https://www.womenintechseo.com/conference", "sameAs":"https://www.facebook.com/groups/698643063902454", "startDate": "2020-03-06", "location": { "@type": "Place", "name": "Barbican Centre", "address": { "@type": "PostalAddress", "addressLocality": "London", "postalCode": "EC2Y8DS" } }, "performer": { "@type": "Person", "name": "Izzi Smith", "worksFor": "Ryte GmbH", "jobTitle": "Technical SEO Analyst" } } } </script>
- 103. 103
- 104. 104
- 105. 105 Entity Knowledge Feature: Powered by: Local Panel & Map Packs Google Discover & News Carousels Product Panels & Surfaces across Google Google Booking Integrations - Google my Business listings - Consistent NAP data - LocalBusiness Markup - Article Markup - Highly trending topics - Published by strong author and domain entity - Product Markup - Product feeds added to Google Merchant Centre - Partner sources - “Hotel ads” - Specific Structured Data
- 107. 107 Nice!But the rankings are useless for performance KPIs
- 108. 108
- 109. 109
- 110. 0.9% of all clicks 53% of all Impressions! https://en.ryte.com/product/search-success/
- 112. 112
- 113. 113
- 114. Your checklist from today onwards! ❏ Incorporate and link to relevant entities within your content to be seen as a reliable source of information ❏ Discover how well your brand is being represented and optimize that for all Search facets. ❏ Structure your Data using JSON-LD implemented Schema.org to be factually present across SERPs. ❏ Exclude entity keywords that you will never earn clicks from, from your KPI reporting and analyses. ❏ Contribute to a better search ecosystem by building more & more strengthened connections.