Lesson 23: Root, Rare, Unique Attributes for Question Generation
Key Takeaways
- So when you look at the root attributes, there is something that you will always be realizing.
- So when you look at the unique attributes, you should understand one thing.
- If you want to create higher relevance, always focus on the unique attributes.
- We always actually protect the proximity between different types of terms as well.
Core Concepts
Main Teaching
Welcome to the 19th lecture. In this lecture, we will talk about the root attributes, rarer attributes, and the unique attributes for question generation methodologies. So when you look at the root attributes, there is something that you will always be realizing. These attributes appear all of the instances of that class.
How It Works
If we choose the city as an instance here, you will realize that city has always a population, always an area, and usually we have a mayor or governor, and you can actually converge it to do something different, but still it is actually a governor. Then we have possibly parks, business centers, game centers, safety-related measurements, and the demographics. These things don't change from city to city, from Berlin to London, from Istanbul to the Hoya or many other cities on this planet. You will actually see these attributes.
Why This Matters
When it comes to the rarer attributes, they don't appear in all the cities, for instance. The city has a nuclear plant, or historical sites, or beaches, forests, or the rivers. When it comes to the unique attributes, you actually have a certain type of feature that can appear only for that city. For instance, Paris actually has the Eiffel Tower.
Implementation Notes
I guess I have written it wrong, but who cares for now? And then we have Sofia for Istanbul. So when you look at the unique attributes, you should understand one thing. If you want to create higher relevance, always focus on the unique attributes.
Koray's Terminology
| Term | Meaning in Context |
|---|---|
| Contextual Bridge | Connections between topical map nodes through hypertext or linkless associations |
| Root Attributes | Attributes that appear in all instances of an entity class (e.g., population for all cities) |
| Unique Attributes | Attributes specific to one entity that act as qualifiers/synonyms (e.g., Eiffel Tower for Paris) |
Practical Application
- Create a content brief using Koray's contextual vector methodology
- 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
- Apply the frameworks discussed to a test website or content network
- Revisit this lesson after completing later lessons to deepen understanding
Connection to Framework
Full Transcript
Welcome to the 19th lecture. In this lecture, we will talk about the root attributes, rarer attributes, and the unique attributes for question generation methodologies. So when you look at the root attributes, there is something that you will always be realizing. These attributes appear all of the instances of that class. If we choose the city as an instance here, you will realize that city has always a population, always an area, and usually we have a mayor or governor, and you can actually converge it to do something different, but still it is actually a governor. Then we have possibly parks, business centers, game centers, safety-related measurements, and the demographics. These things don't change from city to city, from Berlin to London, from Istanbul to the Hoya or many other cities on this planet. You will actually see these attributes. When it comes to the rarer attributes, they don't appear in all the cities, for instance. The city has a nuclear plant, or historical sites, or beaches, forests, or the rivers. When it comes to the unique attributes, you actually have a certain type of feature that can appear only for that city. For instance, Paris actually has the Eiffel Tower. I guess I have written it wrong, but who cares for now? And then we have Sofia for Istanbul. So when you look at the unique attributes, you should understand one thing. If you want to create higher relevance, always focus on the unique attributes. First, then the root, then the rarer attributes. Because when you talk about the unique attribute, it will be a qualifier and definitive measurement or definitive feature for that entity, which will be like using the synonym of that city. It will be like other name of the city. When it comes to the root attributes, you will need to use them for actually providing accuracy and comprehensiveness. Then you will need to talk about the rarer attributes for actually qualifying and signifying that entity in a better way. When it comes to the question generation, mainly we have four types of questions inside our content briefs. One is Boolean questions. Most of the time, they can answer it by telling yes or no. Then we have definitional questions, and usually they start with the question of what. And in this case, actually, we tried to define a certain type of area. Grouping question, usually we have multiple instances from a same type with a context like historical sites, comparative questions, usually we have a superlative or comparative word inside these questions. Of course, there are many other question types that we can use, but most of the time we'll be focusing on these types. And when you look at the Boolean questions inside our actually content briefs, most of the time they won't be appearing at the top. Most of the time they will be actually appearing at the bottom parts like this. And sometimes they will be used for expanding the previous questions context just a little bit further. So if we repeat, every entity has these three type of attribute types, root, rare, unique. And while creating our content briefs, these will be prioritized than this area. I look at the question types, comparative, grouping, definition, and Boolean questions, and we will use these questions with a proper seconds or proper, let's say, order. And while doing that, one more time, we should tell there are different types of other questions too, but mostly we will be focusing on these. At least in this fundamental section, because we will have many different types of lectures as well. So when you look at this specific area too, for defining the Germany, for instance, if we, from this question section or from this query terms, if you come to this area slowly, you will realize that whatever we write at the end of every section, it is a signal or signifier for the next heading. The thing here is that if I... If I search for the H2, whatever we write in the H1 area, it will be a kind of unification of all these H2s. If we have an H3, whatever we write here, there will be a kind of contextual bridge from here to here as well. According to the order of these things, we also create a kind of definitional or contextual opening, then we define and give the answer, then we make a kind of intro to the next heading as well. This way... We always actually protect the proximity between different types of terms as well. So in the next lecture, we will write some questions, then we will order these questions, then we will move in, move on to the hierarchy and the structure. Thank you.