"Web pages are designed for people. For the Semantic Web, we need to look at existing databases"

Tim Berners-Lee’s statement draws a distinction between how information is organized and presented on the traditional web versus how it must be prepared for the Semantic Web. Web pages are crafted primarily with human readers in mind. The content is arranged so people can interpret and understand it through visual layouts, interactive elements, and written language. All the visual cues, formatting, and narrative structures are intended to facilitate human comprehension, navigation, and engagement with the information.

However, this human-centered design poses challenges if machines are expected to process and understand web information autonomously. Computers do not grasp meaning from context, design, or natural language in the same instinctive way as people do. For the Semantic Web to become truly functional, where machines can reason about resources, draw connections, and automate tasks based on content, information needs to be structured in a way that is machine-readable and processable.

Databases serve as a strong example of such structural organization. Unlike web pages, databases store data in highly organized forms, using explicit schemas, standardized formats, relationships (such as tables and fields), and formal semantics. Machines can query, analyze, and draw relationships from database content efficiently because the data representations are unambiguous. By looking at existing databases, researchers and developers can learn how to structure information in a way that allows semantic markup, meaningful, context-aware annotations that are transparent to software agents.

The evolution toward the Semantic Web depends on translating or re-representing information in a way that computers can interpret the meaning, not just the display. Techniques such as Resource Description Framework (RDF), ontologies, and linked data are inspired by principles from databases. The aim is to create a web where information can be linked, processed, and recombined reliably by automated agents, unlocking new possibilities for data discovery and integration, far beyond what is possible with pages solely designed for human consumption.