> ## Documentation Index > Fetch the complete documentation index at: https://docs.antlect.com/llms.txt > Use this file to discover all available pages before exploring further. # Semantic Ontology Layer > How Antlect reconciles the vocabularies used to describe the same capability, so a question reaches the evidence however it was written. Every part of the innovation ecosystem describes itself in its own terms. A government paper records a capability as advanced manufacturing. A university files the same work under mechatronics. A patent classifies it by technical function and uses neither phrase. None of these is incorrect. Each is the natural vocabulary of the institution that produced it. The consequence is that literal search returns whichever fragment matches the words you happened to choose, and the completeness of an answer becomes a function of the questioner's vocabulary. The Semantic Ontology Layer removes that dependency. ## What the ontology does An ontology is a shared model of concepts and the relationships between them. Antlect's ontology reconciles the taxonomies and classification schemes arriving with the source data and maps them onto one conceptual model. * **Terminology resolves.** Different terms for one concept reach the same records, whichever term is used. * **Hierarchy is understood.** A question about a broad field reaches the specific subfields within it, without each having to be named. * **Classifications reconcile.** Patent classification codes, funding programme categories, and research subject codes are read against one another rather than in isolation. ## What it changes about an answer Consider a question about institutional capability in battery chemistry. Matched literally, it returns the records that use that phrase. Interpreted through the ontology, it also reaches work filed under electrochemical energy storage, patents classified by technical function, and grants awarded under a programme category that never mentions batteries. The difference is not a longer list of results. It is that the answer no longer depends on the questioner already knowing every phrase the field uses, which is precisely what the question was asked to establish. ## Where it operates The ontology has no interface of its own. It sits beneath the work: * **[Running queries](/guides/running-queries)** are interpreted through it, which is why questions in plain language work without special syntax. * **[Results](/guides/results)** include records that never used your wording. * **[Graph mode](/guides/network-graph)** groups entities by resolved type rather than by source label. Write queries in the vocabulary natural to you. Reconciling it with the vocabulary of the record is the ontology's work, not yours.