Lesson 6: Introduction to Query Semantics
Key Takeaways
- That's why, despite German and Lithuanian are not that much close to each other, thanks to users, they suddenly become actually similar or relevant.
- So basically, query path is which query has been searched after what query.
- That's why they directly relate these two queries to each other.
- That's why in our topical maps usually we try to go broader so we can always stay safe even if everything goes wrong.
- That's why in the previous lecture we talked about content and the relevance configuration.
Core Concepts
Main Teaching
In this lecture, we will talk about the user behaviors and the query semantics. First, query semantics come from user behaviors and user perception. They don't come from dictionaries. And it means that every word or every search term, they take their relevance and their meaning from the user's behaviors and user's perception.
How It Works
That's why, despite German and Lithuanian are not that much close to each other, thanks to users, they suddenly become actually similar or relevant. And in this case, we will talk about clustering. A search engine can cluster only three things, documents, queries, and users. A search engine's purpose is basically matching queries to topics and entities and topics and entities to the documents and documents to the users so that user can reach out to the document via topics, contacts, and the queries.
Why This Matters
But in this case, sometimes a few components of the search engine can be used to match the queries. And complicated things might happen. Here we have three queries and three users. Let's assume that user one only searches for the query one and query two.
Implementation Notes
Then we see that user three searches for query three and the query two. We see that user two, that person searches for query one and the query three, but the person doesn't search for the query two in this case. So there are a few concepts here that you will need to understand. One is correlative queries.
Koray's Terminology
| Term | Meaning in Context |
|---|---|
| Topical Map | Content network design based on semantics with core and outer sections, processing a central entity with main and minor attributes |
| Topical Coverage | Complete, comprehensive, structured processing of information designed for possible search activities |
| Semantic Distance | The measured gap between two concepts in terms of meaning and query association |
Practical Application
- Analyze the query network for your target topic using search console data
- Create a content brief using Koray's contextual vector methodology
- Map out the central entity and source context for your project
- Study the concepts presented in this lesson until they become intuitive
- Review the related case studies mentioned by Koray for real-world application
- Practice identifying the key terminology in your own SEO projects
Connection to Framework
Full Transcript
Hello. In this lecture, we will talk about the user behaviors and the query semantics. First, query semantics come from user behaviors and user perception. They don't come from dictionaries. And it means that every word or every search term, they take their relevance and their meaning from the user's behaviors and user's perception. That's why, despite German and Lithuanian are not that much close to each other, thanks to users, they suddenly become actually similar or relevant. And in this case, we will talk about clustering. A search engine can cluster only three things, documents, queries, and users. A search engine's purpose is basically matching queries to topics and entities and topics and entities to the documents and documents to the users so that user can reach out to the document via topics, contacts, and the queries. But in this case, sometimes a few components of the search engine can be used to match the queries. And complicated things might happen. Here we have three queries and three users. Let's assume that user one only searches for the query one and query two. Then we see that user three searches for query three and the query two. We see that user two, that person searches for query one and the query three, but the person doesn't search for the query two in this case. So there are a few concepts here that you will need to understand. One is correlative queries. Query path and sequential queries. So basically, query path is which query has been searched after what query. If query one has been searched first, then the query three, it means that query path goes like query one and query three. If the person searches first, let's say the query two, then the query one, it goes two to the one. The second thing is correlative queries, which query has been searched together with another, which another. In this case, we see that actually none of these three queries are being searched together. Only two of them always are being searched together, which means that there is not an absolute correlation between them. Then we see the sequential queries. Sequential query is about actually a kinds of search session. Let's say this line here, it represents a search session and this line here, it represents another search session. In this case, you can see that the query three maybe appears in this part and query one appears in this part. The search engine decides that both of these query sessions appear from the same search context or the user state. They don't take this middle ground and they directly assume that these two queries are actually coming from the same search state. That's why they directly relate these two queries to each other. But the thing here is that how we can create a query cluster because none of these three things are being actually searched at the same time. If you accept that these three users are actually the same user, then you will realize that the same user actually are searching all these three queries at the same time. The real magic happens actually where you put the border because if you take these two users, then the query paths and the correlations and the sequential queries, they all will be changing. If you take it like this again, everything will be changing one more time. If you take it like all these three, then it's easier. But this time you can realize that some of the really small websites, small sources, they might lose traffic. Or some of the affiliate sources, suddenly their topical coverage might be decreased too. Because if you are targeting these users here, actually their behaviors and their entities that appear inside these queries and the query context, all these things actually will be determining your borders. That's why in our topical maps usually we try to go broader so we can always stay safe even if everything goes wrong. One more thing that you should keep in your mind. I won't talk about the theories that much in the future lectures. Is actually semantic distance. In every code algorithm update, the semantic distance between the topics, it changes. That's why in the previous lecture we talked about content and the relevance configuration. It is an always-on process. And that's why in our content briefs, we never leave a gap. We cover everything without any kinds of luck. We determine everything. Nothing is random. Everything is planned. And we cover everything in the best possible detailed and highly accurate way. So that Search Engine, can't find any chance besides ranking us. Thank you. In the next lecture, we will see the topical map of this project.