Using AI and Knowledge Graphs to Maximize Content

How to use AI and Knowledge Graphs to Maximize Content

Knowledge panels are hard-won and highly coveted, yet many SEOs and content writers find them mysterious aspects of organic search.

Today, we know that large language models (LLMs) are still in early development; and that SEO schema markup is ideal for training machine learning. From participating in Google events and reading patent applications, most have gained a basic understanding of Google Knowledge Graph (KG) entities. This helps SEOs incorporate knowledge graphs and knowledge panels as a mature, advanced strategy for long-term and mission-critical use cases.

Both business brand entities and personal and want to secure their own richly-filled and accurate Knowledge Panel entry. It can be influenced by using Organization and People Profile Page schema. This article is about how we can “influence” them, even when what displays in them is out of our “control.”

Table of Contents

Logically, before you can get your content display in a KG, you need one.

How to Gain a Google Knowledge Panel for Your Business?

Google tells us that knowledge panels are automatically generated for searchers conducting queries for entities that are in the Knowledge Graph. Its panels use Artificial Intelligence (AI) to automatically update “as information changes on the web.”

Conduct a Google Search for your business name and then click on “Own this business?” By linking in the panel, you can claim it. The verification process is straightforward – if you can prove ownership. Google can easily associate the knowledge panel entity through one of the following accounts; Google Search Console, Looker Studio, YouTube, Twitter, or Facebook. If you aren’t a known entity there, then you may be required to provide additional information that confirms you are authorized to claim it.

Your personal or business knowledge panel provides a summary of information related to the search query, including a brief description, images, related entities, and more – about you! Inaccuracies need skillful KG management.

How does Google Determine which Entity Data to Display?

We can build on the foundational understanding that Google pulls information from its Knowledge Graph for knowledge panels. Next, we can learn a bit about how AI gleans information for the Knowledge Graph.

The SGE will likely radically change once it is fully rolled out. However, we know that it seeks to offer a combined response from different, trusted resources into text that provides a comprehensive text answer to the inquiries.

While traditionally, Wikipedia was most frequently the source of information, today we’re seeing that expand. With the emergence of Google’s Search Generative Experience, our tests indicate that Google can use information from any crawlable published source for their Knowledge Graph.

The types of sources that Google tends to include in its SGE query responses are:

  • Accurate.
  • Use expert authors.
  • Have a following.
  • Structured.
  • Consistent.

This presents a clear need to engage in the responsible use of AI since it possesses unique capabilities that can be exploited or misused. Experts are in demand who know how to use it as a content-enhanced productivity tool.

Bing and Google Search want to give people quick access to information in knowledge panels through a simplified natural language-based information retrieval process using AI.

How do Knowledge Graphs Use AI?

It is now critical to learn how to connect the dots of your data through advanced knowledge graphs

In the AI space, knowledge graphs complement machine learning techniques in multiple ways. They lessen reliance on large, labelled datasets; help to transfer learning and explainability; encode domain, task and application knowledge versus the costly processes of learning from data alone.

For example, formerly Google has changed it HowTo and FAQ schema markup reliance to help match queries with answers. Today, that has changed as the tech giant’s AI capacity can surface less-cluttered SERPs without relying on it as heavily.

AI is useful for gleaning insights for graph management. It can help in educating LLMs and assessing their capabilities for specific content management tasks. LLMs can can reduce time and current costs.

AI and LLMs can democratize knowledge graph usage for:

  • RAG and NLQ: Retrieval-augmented generation (RAG) enables data communication to get instant insights; natural language query (NLQ) helps you ask questions of your data using business terms. This can be done without writing SPARQL.
  • Natural Language Processing: converting unstructured text into structured data while detecting new types of semantic relationships.
  • Essential data management: Classifying business documents to empower precision search

For brick and mortar stores selling in person to customers, local business schema is helpful for data management.

How to Build your Knowledge Graph to Expand Content Reach

  1. Establish a solid entity home.
  2. Leverage Large Language Models.
  3. Implement schema markup.
  4. Implement the concepts of topic authority.
  5. Be agile with search change but grounded in solid content strategies.

It can be complex to know how to optimize your content strategy for Google SERP features in its new Search Generative Experience. When striking out into new SEO techniques, it helps to learn from a front runner’s testings. So, I offer a summary of mine here.

1. Establish a solid entity home

The home of your entity is where you clearly state who you are and what you do. Some use a robust social profile, a personal website, a brand’s “About us” page, an author bio page, or wherever exists online an accurate and recognized source of information.

Your entity home should contain at least one prominent and clear sentence describing who you are and what you do. This sentence should then be added to all bio information about your personal or brand entity published on the web. Add outbound links to sources that accurately cover you or your company. It is important to use trustworthy sources.

Google looks for verification and consistency when determining how trustworthy you are for inclusion. Meaning, that confusing or conflicting descriptions may delay your KG ownership and getting your content into Google’s question answer SERPs.

Updates to your Google Knowledge Panel may be derived directly from the entity in the panel or general user feedback. This is why you need someone to keep a continual eye on it. It may update at any time due to someone else’s input.

2. Leverage Large Language Models

Knowledge Graphs and Large Language Models go hand in hand. The rapid adoption of ChatGPT and emerging LLMs have already vastly impacted content creation workflows and the field of knowledge graphs. With prompt refinements, they can generate logical and natural language text.

Knowledge graphs maintain multiple benefits over LLMs. Most importantly, knowledge graphs consist largely of structured data. Actually, knowledge graphs can store both structured and unstructured information. This assists search engines when facing complex queries, and the need for information extraction demands quickness and accuracy.

While improving, a current weakness of LLMs is that they are black-box models and often lack factual knowledge. One of Google KGs’ greatest strengths is that it looks for factual credibility and Experience, Expertise, Authoritativeness, and Trustworthiness (E-E-A-T) before adding sources to it. KGs are, essentially, collections of facts, and they are fully interpretable.

3. Implement schema markup

Graph structured knowledge: JSON-LD, RDF, and schema.org are key to KG management.

Well-planned schema markup brings your content alive in the world of robots needing to retrieve information quickly. We almost always use Organization schema markup to confirm that same information in Google’s native digital language. It can highlight your key points, authors, how-to pages, factual statements, product details, customer reviews, and so much more. It helps search engines feel confident about your brand and what you are saying.

We focus a lot on schema markup that lets us leverage SameAs data. It is a helpful way to include linked entities that confirm relationships and skills.  Use the specific schema type that most accurately applies to your context.

Choice selection of structured data markup implementation provides search engines with a contextual understanding. In turn, this may increase visibility in many forms, but especially question answer content in your Knowledge Graph.

We help clients who face challenges when seeking knowledge about the different restrictions for schema production at scale. It’s easy to find yourself battling schema markup drift when managing multiple page types and edge-case scenarios.

4. Implement the concepts of topic authority

We know that internal and external links provide trust signals. However, how your web of data is arranged matters, too.

There are several ways to manage topic authority, such as a collections page, the use of entity tagging, and UX hierarchy. How you manage this on your website, helps both people and search engines recognize your expertise, services, products, and helpful content.

By linking to your entity home within external publications, conversations, podcasts, etc., you are helping to reinforce your entity home and E-E-A-T signals.

Content practitioners that adhere to such content strategies can best achieve this by including some level of AI in their everyday workflow. Machine learning algorithms are powering search; you will understand them better if your content categorization and tag management can “think” like they do on an entity level.

You may win visibility in Google’s “People Also Search For” feature and the new SGE “Ask a follow up” question feature when you’re recognized as an expert.

5. Be agile with search change but grounded in solid content strategies

Trends and fads come and go, but the SGE is here to stay. We hear this directly from Google. This implies the necessity to adapt where needed and keep mainline content marketing strategies in place. This is where content consistency is a key factor.

Obtaining high content reach per article requires aligning content efforts with business priorities and across teams. Each publication also benefits from robust article schema structured data. Your internal staff and freelance marketers need to align on what you’re seeking to accomplish and how to do it. Outcomes and measurements should be clear so that they offer the best insights and are correctly implemented.

By aligning on a smart, doable content strategy, it will be less work for more gain.

With the right strategy to thread your entity relationships within your content, you may first gain a handful of content pieces in your KG. From there, it will make it easier for article authors to write pieces with high appeal to your audiences and establish a more robust KG. Focus on evergreen green, upper funnel content to increase your conversions.

“Good content” is becoming “average content” with how easy AI makes content generation. By humanizing your efforts, you add the finishing touch that connects with people. Add depth to content by using structured data (and knowledge graphs) that reduce friction for your content authors.

Updates to Google search algorithms are continuous. For example, already in August, Google announced that searchers can now “Get a better understanding through images and videos. Sometimes, it’s more powerful to understand something by seeing it — so we recently brought images to even more AI-powered overviews.”

AI Populated Knowledge Graphs Make Search More Effective

For most of us, this change is occurring faster than we can keep pace. Yet, it can also make us more effective in our approach to content creation. From mobile content modeling to conversational content design, by embracing a dialogic framework and AI hints and refinements, you can leap ahead in competitive markets.

Google offers advice for content creators who are ready to try AI-generation. Its verbiage has changed as its SGE has emerged.

Think of KG AI as replicating how text auto-fill works. LLMs are constructing conversations that may be challenging for humans to grasp because we think linearly. We never create content for AI bots, but we can leverage AI bots to refine our content creation to reach people.

This technology is relearning as it goes. It seeks to mimic our human cognitive processes. Rather than spinning out mediocre AI-created content, you have more pressure to generate quality content.

“The pioneers of AI research recognized the potential benefits of AI and humans working together and envisioned a symbiotic relationship between the two. They anticipated AI would augment human intelligence to solve complex problems and make better decisions and ultimately help humans achieve their goals more efficiently”

 

The main deficiency is that you have no idea where the knowledge in the answers come from? Thus, you cannot judge or evaluate the quality of the answers. The other important concluding point is that the more pre-existing knowledge you have the better you can use ChatGPT and ensure you get quality information out of the use.” – International Journal of Information Management: So what if ChatGPT wrote it?, August 2023

Super Helpful Content that Engages Your Active Audience

It is not about “producing content”; it is about “connecting through content”.

Digital communications must adapt to users’ unique communication scenarios around the same truths. How your content is framed and communicated should be tailored to the moment and listener. Technology should not be the determinate factor; rather, what works for creating human connection should be. You cannot own a competitive differentiation factor via figures, KPIs, etc., alone. You can establish differentiation by touching someone’s heart.

Clarify your messaging and writings to offer the most value. The reason it takes me a while to publish a new content piece is that I “think” it through. When rushed by deadlines, deliverables, content calendars, or whatever, you may publish less than your best content. And AI Generative content needs a skilled, human”thinking” review to determine if your content is truly customer-centric.

It’s About Creating Human-Touch Quality Content

Imperfections exist in both human-generated and AI-generated content. Whether content is human expert-produced or AI-produced, Google’s systems are going after high-quality information from reliable sources. It states, “not information that contradicts well-established consensus on important topics. On topics where information quality is critically important—like in healthcare, civic, or financial information—our systems place an even greater emphasis on signals of reliability.”

We interpret this to indicate that the future of SEO lies in the balance between embracing the benefits of ChatGPT and maintaining the quality of publications. The human touch is vital even though your Knowledge panel and knowledge graph consist of AI-populated content.

As explained, however, content is produced, those seeking success in Google Search should be looking to produce original, high-quality, people-first content demonstrating qualities E-E-A-T.

 

Creators can learn more about the concept of E-E-A-T on our Creating helpful, reliable, people-first content help page. In addition, we’ve updated that page with some guidance about thinking in terms of Who, How, and Why in relation to how content is produced.

 

Evaluating your content in this way, whether you’re using AI-generated content or not, will help you stay on course with what our systems seek to reward.” – Google Search’s guidance about AI-generated content

By planning for featured snippet opportunities, your web content may be pulled by AI generative search when a query relates to yourself or your business. This is a way to increase visibility and influence your own knowledge panel. When your target audience sees you show up alongside a featured snippet, it demonstrates your authoritativeness and ability to be helpful.

Embrace time-saving AI processes

A better understanding of the intertwinings of AI answers and knowledge graphs can significantly help expand the reach of your content marketing efforts. It’s an exciting time to be a content marketer. It requires the ability to skillfully chain your prompts for ChatGPT or Google Bard. The more granular your prompts are, the better.

The following tips will help your AI output be a game changer:

  1. Create a resource library to inject quality data.
  2. Instruct your AI tool step by step.
  3. LLMs have limited knowledge and are not an expert on everything.
  4. Retrieval-Augmented Generation (RAG) helps control long-term memory and these LLMs model’s knowledge and accuracy.
  5. An information retrieval-based memory helps improve response accuracy.
  6. Add your own sourced, factual support details; don’t rely on AI for factual content.

The stakes of using Generative AI to write content are high. But dig in! In the long run, really watch engagement levels with your content. Being human and subjective in your content remains critical. You should trust AI as much as you trust human opinion. Ask yourself, who is the governor of truth? An expert’s opinion, experience, and learnings vary by person. This is why Google says that “trust” is the most important element in E-E-A-T.

Content creation’s future needs experts who can build multi-agent systems capable of extracting in-context data using AI-built knowledge graphs for audience engagement.

Content creation’s future needs experts who can build multi-agent systems capable of extracting in-context data using AI-built knowledge graphs for audience engagement.

The stakes of using Generative AI to write content are high. But dig in! In the long run, really watch engagement levels with your content. Being human and subjective in your content remains critical. You should trust AI as much as you trust human opinion. Ask yourself who is the governor of truth? An expert’s opinion, experience, and learnings vary by person. This is why Google says that “trust” is the most important element in E-E-A-T.

Content creation’s future needs experts who know how to build multi-agent systems capable of extracting in-context data using AI-built knowledge graphs for audience engagement.

Google and others offer advice for content creators ready to try AI-generation.

“As AI-powered virtual assistants like Siri, Alexa, and Google Assistant become more integrated into our daily lives, optimizing content for these agents is crucial. Traditional SEO focuses on improving website visibility on search engines. In contrast, AI Agent Optimization (AIAO) is about tailoring information, products, and services to be easily discoverable and consumable by AI assistants.

 

If conversational flow is the script and the AI agent the actor, then the data structure and schema act as the stage setting: foundational, supportive, and critical for the play to unfold.” – Javier Calderon Jr [1]

“What remains in a reader’s emotional memory is a lot more than what remains in the cognitive memory.

 

Set aside the tech speak we are used to, not productizing what we do, but rather relating well through the craft of marketing communications. – @TheodoraPetkova with @jasonmbarnard. The conception work is “thinking” and exchanging versus being ‘metadata focused’ so much. How are you going to differentiate your craft in the midst of AI sleekness? – Teodora Petkova [2]

“By definition (I’m simplifying), if you’re using AI to write your content, it’s going to be rehashed from other sites.” – Google’s John Mueller [3]

“By definition (I’m simplifying), if you’re using AI to write your content, it’s going to be rehashed from other sites.” – Google’s John Mueller [3]

SUMMARY: Use AI to Influence Your Own Knowledge Panel

We believe it is best to rise about fear of chance and the negative chatter – because how people find and consume content is changing! The rising level of content consumption is staggering. While some craft AI content with little thought and integration, those who leverage it wisely may win on the SERPs. Since people use multiple screens simultaneously, you need a solid approach to connect your brand with your intended audience.

Call Hill Web Marketing to gain a strategic approach to help you stand out and drive revenue impact. 651-206-2410

To learn how Google views your organization, Request a SERP Analysis

 

RESOURCES

[1] https://xthemadgenius.medium.com/ai-agent-optimization-the-next-frontier-beyond-seo-bdfa25a3264d

[2] https://www.youtube.com/watch?v=yElS16E0oCg

[3] https://www.seroundtable.com/google-ai-write-content-rehashed-35884.html

Jeannie Hill:

This website uses cookies.