Topics
12 tags, assigned at index time. Less precise than a vertical, and they work across verticals — useful when the same subject appears in news, documentation and academic writing.
Technology
technologySoftware, hardware, infrastructure and the industry around them — assigned from what a page is about, so it holds across news, docs and papers.
Science
scienceResearch findings, methods and scientific reporting across disciplines, from primary papers through to the coverage that summarises them.
Health
healthMedicine, public health, clinical research and health policy — clinical material together with the policy and reporting that surrounds it.
Finance
financeMarkets, banking, monetary policy and corporate finance, including the regulatory material that governs each of them.
Business
businessCompanies, strategy, funding, operations and industry structure — the organisational side of a subject rather than its technology.
Politics
politicsGovernment, policy, elections and regulation, from primary legislative and regulatory texts to the reporting and analysis around them.
Education
educationTeaching, learning, institutions and educational policy, across schools, universities, vocational training and professional development.
Sports
sportsCompetition, results and the business of sport — fixtures and match reporting alongside governance, transfers and the commercial side.
Entertainment
entertainmentFilm, television, music, games and the industries behind them, covering both the works themselves and the business of making them.
Food
foodCooking, ingredients, restaurants and food production, from recipes and technique through to agriculture and the supply chain behind them.
Travel
travelDestinations, transport, accommodation and travel logistics, including the visa, safety and regulatory material that constrains them.
Lifestyle
lifestylePersonal finance, home, wellness, relationships and consumer advice — the practical, personal end of a subject rather than the institutional one.
A closed set, on purpose
Topic tags are a controlled vocabulary rather than free text, which is the whole reason they are useful to an agent. A free-text subject field forces the agent to guess the string that will match, and a guess that misses returns nothing without saying why. A closed set can be read: /describe returns the current list unauthenticated, so an agent can discover the valid values instead of carrying a copy that goes stale.
The tags are assigned at index time from what a page is about, not from the domain that hosts it or the section of the site it sits in. That is why a topic holds across content types — the same subject is found in news reporting, in documentation and in a paper, which a domain-based rule would never have connected.
Using them well
A topic is a coarse instrument and works best as one half of a pair. On its own it narrows the corpus; combined with acontent type it describes a shape of document, which is usually what the question actually wanted — technology plus documentation is a very different corpus from technology plus opinion, and only one of them will answer a configuration question.
The boundaries between topics are soft by construction and a document can carry more than one. Treat a tag as an inclusion hint rather than a partition: setting one narrows what comes back, it does not guarantee that everything relevant carried that tag. When breadth matters more than precision, leave the topic off and use a quality floor instead.
Find the authoritative sources on a topic
The authorities endpoint ranks by graph centrality rather than keyword match — what the field treats as foundational, not what matched your words.