Semantic SEO Glossary: Complete Guide to Key Concepts

Topical Authority

"Topical Authority is ranking over an authoritative website for a certain amount of time with a lower cost-of-retrieval, higher accuracy, clarity, and information responsiveness by creating semantically organized content networks in the form of main and supplementary content by optimizing micro-macro semantics and contexts."

Formula: Topical Authority = Topical Coverage × Historical Data

This definition requires understanding main content, supplementary content, content configuration, micro-macro semantics, macro-micro contexts, contextual domains, knowledge domains, EAV, and semantic content networks.

Historical Data

A critical component of Topical Authority. Historical Data is not about "time"—it's about "user engagement" and the quality of that engagement.

• Includes mouse-overs, impressions, even a ranking in the 94th position
• Non-quality user engagement or query session logs can demote your website after a certain amount of time
• If you lose rankings today, it's because of historical data from 6 months ago
• Cleaning bad historical data requires good historical data with a stronger signal

Topical Coverage

Topical Coverage is NOT measured by the amount of web pages or mentions of entities. Stuffing entities and attributes won't provide higher coverage.

• If you didn't define X, you didn't cover it
• If you didn't connect X to Y, you didn't cover it
• If you didn't match macro-context to query context, you didn't cover the entity with proper context
• If your definition misses certain aspects, you didn't cover it

Topical Coverage is determined by the complete-comprehensive and structured process of information on web documents designed for possible and related search activities.

Relevance

Relevance and Responsiveness are NOT the same.

Relevance is for improving the Information Retrieval Score, which comes from term-weight calculation and reading the retriever's mind with math algorithms.

For relevance, the web document must be relevant to all possible interpretations of the query.

Responsiveness

The direct Information Extraction Process—requiring a direct answer that satisfies possible or related search activities.

To provide responsiveness, the web document must satisfy all possible needs behind the query, not just be topically relevant.

Query Semantics

The meaning and intent behind queries as understood by search engines. For single-word queries like "cancer," the distribution of probabilities comes from query variations, search behaviors, and query search demand.

Search engines interpret queries not just based on query semantics—sometimes they assume the highest PageRank or most authoritative source interpreted the query correctly and bring similar web documents to the SERP.

Macro Semantics

The overall semantic characteristics of a content network from the point of view of semantics. Macro semantics consider:

• Site-wide N-grams that appear on the website
• Most used nouns, adjectives, and predicates site-wide
• Most used question formats site-wide
• Types of queries targeted overall
• Heading vector construction
• First words of paragraphs—do they answer the question or use rhetoric without information?
• Heaviest context terms in web documents

Micro Semantics

Word-by-word optimization of documents for higher relevance and responsiveness. Involves sequence modeling and optimization of visuals with subject and object entities.

• Subject Entity: The most concrete and prominent object in visuals
• Object Entities: Interpret the context of the image by changing certain interpretations
• Sequence Modeling: Changing word sequences for higher responsiveness (e.g., "Teacher yelled students" vs "Students are yelled by Teacher" distribute relevance differently)
• Predicates signal overall context—"yell" annotates a different context from "shout"

Main Content

The main part of a web document that processes the macro-context.

• Involves all context-terms, topical entries, and main entities
• Gives the main context-flow and coverage with a summary of the entire article
• Doesn't touch sub-contexts or minor topics
• Provides proper connection to query contexts and semantics
• Doesn't involve too many internal links unless inside the macro-context

Supplementary Content

Content that touches micro-contexts, sub-topics, and provides internal links to side-topics.

• Provides better association and "Neighborhood Content" between different topical map segments
• Always connected to macro-context but processes it with connection to another macro-context

Topical Map

A concept and phrase first used by Koray Tugberk GÜBÜR during the interview with Matt Diggity. To create a proper topical map, you must understand:

• Source Context – The purpose and monetization strategy
• Central Entity – The entity appearing in every section
• Central Search Intent – Source context + central entity unified
• Core Section – Main attributes, densified per source context
• Outer Section – Minor attributes, trust/quality signal propagation

A topical map is NOT just "connections between entities via attributes" or "blog post names without understanding context or semantics."

Source Context

The purpose of the source (website and web entity including CEO, social media platforms, and collections of other web aspects of the brand), and how the brand monetizes its content to turn search engine users into customers.

• Must be connected to the Central Entity with a proper attribute
• Reflected site-wide through N-grams in every web page, boilerplate, and main content
• Determines how related topics connect (e.g., "climate" connects to "Real Estate Investment" for real estate consultancy, not to "best time to visit")

Core Section (of Topical Map)

The unification of the source context with the central search intent. The core section focuses on a specific main attribute of the central entity.

• The main attribute comes from the source context
• For affiliates, the "quality" attribute with derived attributes like "durability, charge time, maintenance"
• For engineering companies, the "production" attribute with derived attributes like "materials, designs, types"
• Must be densified according to the source context

Outer Section (of Topical Map)

The section of the topical map designed to improve overall historical data and increase overall topical relevance and contextual consolidation.

• Focuses on minor attributes of the entity, not main attributes
• Propagates trust and quality signals to the core section with links or linkless connections
• For "Visa Consultancy": Core focuses on "Visa," outer focuses on country attributes like "religion" or "language schools"
• For "Divorce Lawyer": Core handles divorce/legal predicates, outer covers marriage dynamics, roles, and functions

Central Entity

The entity that appears in every sub-section of the semantic content network, whether in main content and macro context, or supplementary content and micro context.

• Gives its main and minor attributes to the core and outer sections of the topical map
• Always appears inside anchor texts with synonym value
• United with Source Context to create connection to users' possible and related search activities
• Determines the direction of the topical map for proper classification with authoritative sources

Central Search Intent

The intent that appears in all of the topical map and semantic content networks, whether in boilerplate or main content. It is the unification of the source context with central entity.

• Heavily processed in the core section of topical map
• Reflected on the outer section of the topical map
• Guides the overall content strategy and contextual focus

Entity-Attribute-Value (EAV)

A data model structure fundamental to semantic SEO where entities are described through their attributes and corresponding values. Understanding EAV is essential for creating semantically complete content.

Stuffing entities or attributes won't provide higher rankings—instead, you must define concepts, connect them properly, and match macro-context to query context.

Content Configuration

The process of changing and updating existing content according to changed semantic distances or similarities, and increasing relevance and responsiveness continuously.

This is an ongoing optimization process that responds to shifts in how search engines interpret queries and relationships between entities.

Publication Frequency (Momentum)

The frequency of major content updates or new content publications to take the attention of the search engine to be prioritized for being crawled, indexed, and ranked earlier and higher.

Part of the Vastness-Depth-Momentum triad—if you can't go faster, create much more wider and deeper content.

Contextual Coverage

The context that covers a certain portion of a web page.

• If a context is heavily and vastly processed, it will dilute the prominence of other sections
• If a context is processed lightly but connected to certain query interpretations, it will decrease relevance

Contextual Flow

The order in which contexts are processed on a web page. Processing the same things with different order creates different possible click satisfaction scores.

• Heading formats, words, and hierarchies change context priority
• Order of information presentation impacts semantic interpretation

Contextual Hierarchy

The adjustment of a context's coverage weight. Represented with typography, visuals, and coverage of specific sub-sections of the web page.

• Changes the weight of an internal link for relevance and PageRank pass
• Changes the macro-context of the web page
• Critical for proper semantic signaling

Contextual Border

The border between macro-micro context sections of the web page.

• Provides a slow transition from main content to supplementary content
• Provides a grouper question to deepen the main context while connecting it to other side-topics

Contextual Bridge

The connection between two different topical map nodes.

• Provided by aligning and consistent information without a link
• Also provided with a hypertext for contextual connection between different topical map nodes

Vastness-Depth-Momentum

A simplification principle: Go wider, go deeper, go faster.

• If you can't create a "wide source," go even deeper and faster
• If you can't go faster, create way more wider and deeper content
• Whichever is missing, complete its effect by improving another
• Create a bigger topical map OR really deep content briefs for a small topical map
• Publish lots of things quickly OR publish deep, comprehensive documents slowly

Semantic Content Network

A collection of connected web documents that are semantically optimized for a topic. The quality of a semantic content network can affect the rankings of the web source for other topics as well.

Non-quality pages within the network will make other quality pages rank lower.

Semantic Distance

The conceptual distance between queries and content. Semantic distance changes continuously, meaning during and after BCAUs (Broad Core Algorithm Updates), SERP characteristics and ranking web source types change tremendously.

Semantic distance is measured using two main methods: calculating association and connection angles between entities, or counting the length of the connection between things with certain associations.

Semantic SEO

Semantic SEO is the process of creating a content network in a relevant and meaningful structure for each entity within a subject. It involves connecting terms, entities, and facts to each other with factual accuracy and relational relevance.

• Focuses on meanings and topics instead of words
• Aims to better satisfy user search intent
• Establishes authority for both Search Engines and Users
• Creates organized content structures with clearly connected entities

Entity

An entity refers to a thing or concept that is singular, unique, well-defined and distinguishable. Entities can be a person, place, object, or even an idea.

They play a crucial role in understanding and processing natural language data, and in generating more relevant and logical content for Search Engine Results Pages.

Triple

A Triple consists of an entity, predicate, and object. It is used in NLP and by search engines for organizing information on the web.

• Google uses triples for indexing entities and re-organizing web documents
• Example: "Tom Hanks played" or "Tom Hanks said" types of queries
• Entity-Attribute-Value (EAV) is another type of triple in semantic SEO

Semantic Similarity

Semantic Similarity refers to the closeness and relevance between two words. It involves:

• Lexical semantics: Semantic relations between words
• Semantic closeness: The distance between words' meanings

Semantic Relevance

Semantic Relevance refers to the degree to which a certain term or concept is related to the given context or topic. It's about how closely the meanings of two entities align with each other in a specific context.

Information Responsiveness

Information responsiveness means being responsive to all related and possible search activities in every form of the search query with all context interpretations as a web source.

• Requires query-question-answer pairing and indexing
• Connected to Passage Indexing (Ranking)
• The web document must satisfy all possible needs behind the query

Relevance for Information Retrieval

The score from text processing methodologies such as term saturation, length normalization, co-occurrence matrix construction, BM25, TF-IDF, GloVe, Word2Vec and more.

It shows the overall connection between the query and information available, though it's considered the "Blind Librarian" state as it focuses on connection rather than direct answers.

Query Network

A Query Network represents a search language with all possible word distributions by representing terms in queries with different aspects, definitions, and themes:

• Query Aspect: An angle for the search term with context signifiers (e.g., "Hawaii Hotels" represents the "aspect of holiday")
• Query Definition: The phrasified version of query terms (e.g., "Reserve a room in Hawaii Hotel")
• Query Theme: The direction of the query aspect, matched with the web document's theme

Represented and Representative Queries

Concepts from Query Processing where a search term has both representative and represented query versions.

• Representative Queries: Broad (e.g., "Healthy diet recipes")
• Represented Queries: Specific (e.g., "Healthy dinner recipes for weight loss")
• The representative query has access to the represented query's relevance
• "board vision" and "vision board" give different results due to different contextual connections

Knowledge Domain

A Knowledge Domain refers to specific areas of queries, entities, layout designs, search patterns, and user segments with specific information structure and user-satisfaction models.

• Contains specific information, layout design, and sentence-information structure
• Search engines classify publishers based on specialty, authority, and coverage
• Includes "experts", "apprentices", and "laypersons" based on expertise level
• Example: In Currency Domain, high bounce rate can be positive as users check rates and leave

Contextual Domain

A Contextual Domain refers to the specific area of a query where search intent, user behavior, and content structure are defined according to specific context.

It impacts how search engines interpret and rank content, and influences user behavior and content interaction patterns.

Contextual Layer

A contextual layer represents a specific level of depth and detail within a contextual domain. Characterized by processing angles and context qualifiers.

Contextual Vector

A vocabulary list created with macro-context for each unique term from a domain, based on term occurrences. Part of Google's User-context-based Search Engine Patent.

Helps Google understand unique aspects within a domain, differentiating user behaviors, expectations, and quality parameters.

Natural Language Processing (NLP)

A fundamental AI subset that allows computers to have meaningful discourse with humans using natural language. NLP employs Machine Learning, computational linguistics, and statistical analysis techniques.

Sliding-window in NLP

A concept explaining the width of tokenized and processed text. Used by neural networks and LLMs to predict next words, sentences, and syntactic/semantic meanings.

Example: "Koray bought stocks from NVIDIA by trusting the cost of LLM training" — with a 3-word window, first iteration only processes "Koray bought stocks."

Sequence Modeling in NLP

Changing word sequences for higher responsiveness and contextualization. A type of statistical modeling for predicting the next word based on preceding words.

Example: "Teacher yelled students" vs "Students are yelled by Teacher" — these distribute relevance differently!

Root Document

A central, independent element (page) in a topical map that doesn't primarily focus on the outer section. Acts as a hub linking to the most significant elements.

• The H1 heading is the most significant heading in the entire topical map
• H1 is directly connected to the Central Search Intent
• Most important attributes linked at top, less important at bottom
• H1 heading should be linked to other documents in the network

Node Document

A web document that forms part of a semantic content network. Two types exist:

• Quality Node: Created to achieve high search rankings
• Non-quality Node: Covers secondary aspects, not intended to rank highly

Semantic Content Brief

A document containing comprehensive information for creating a Semantic Content Network. Includes essential details related to lexical semantics, entity relationships, and relevant phrases.

May utilize phrase-based indexing, word vectors, or context vectors to assess contextual relevance within a specific contextual domain.

Title (Meta Title)

The Meta Title of a webpage, displayed on search engine results pages. Provides a brief and accurate summary of webpage content. Google may use H1 Tags instead if they better define the webpage's purpose.

Description (Meta Description)

A brief summary of the webpage's content, function, and purpose appearing on SERP. Can be a small direct ranking factor. Should not exceed 320 characters to avoid being shortened by Google.

Image ALT Tags

Used to mark images within HTML documents. Crucial for SEO as it helps Google understand image content.

• Should describe image content and purpose on the webpage
• Increase relevance for user-intent and search intent
• Important for web-accessibility
• Should be created considering visual search on Search Engines

Attribute Prominence

How essential a specific attribute is to an entity's definition.

Attribute Popularity

The frequency and volume of searches related to a specific attribute of an entity. Measures how often people search for information about this attribute.

Example: If 'smartphone camera quality' receives high search volume, it indicates camera quality is a popular attribute.

Attribute Relevance

How relevant a particular attribute is not to the entity itself, but to the Source Context. This determines which attributes should be prioritized based on your website's purpose.

Lexical Semantics

The study of the relationship of words to each other. In lexical semantics, you focus on how close or distant words are according to their meanings—going beyond keywords to understand word relationships.

Holonym

A term that denotes a whole, the part of which is denoted by another term.

Example: "Face" is a holonym of "eyes"

Meronym

A part of the holonym. By combining with other meronyms, it makes up the holonym.

Example: "Nose" is a meronym — combined with "eyes", "lips", etc., they make up the "face" (holonym)

Hypernym

A broad term for a group.

Example: "Color" is a hypernym of "red," "green," and "yellow"

Hyponym

A specific term that comes under a general term or group.

Example: "Red" is a hyponym of "color"

Synonym

Words with different pronunciation or spelling but the same meaning.

Example: Hard, difficult, challenging, and tough are synonyms

Antonym

Words that are related but have opposite meanings.

Example: "Sleep" and "Awake" are antonyms

Polyseme

Words that have two or more related meanings.

Example: "Bright" can refer to "shine" and "intelligence" simultaneously

Homonym

Words that have the same spelling and pronunciation but different meanings.

Example: "Bear" — as a noun (animal) vs as a verb (to endure)

Skip-gram Dominant Words

Identified through word embedding models like Word2Vec. Represent words that frequently co-occur with many other words in a corpus.

• Act as anchor points in the embedding space
• Shape relationships between other words
• Useful for word sense disambiguation, summarization, and topic modeling