Semantic technology is a concept in computer science that aims to bring semantics — the meaning and context behind words and sentences — to the world of computers. A number of approaches to implementing the concept have been developed, ranging from advanced artificial intelligence to formal, machine-readable descriptions of content. The Web is a key focal point for semantic technology, though it may benefit business and academic fields as well.
Although computers excel at mathematical calculations, they struggle with many aspects of human language, especially semantics. A computer program can defeat even the most skilled humans in a game of chess, but would fare poorly in a trivia contest against a child because it lacks the ability to accurately interpret the context, meaning, and subtleties of the language in the trivia questions. This has implications for a great range of applications and services: Without a thorough understanding of context, a search engine may not return accurate results for words with multiple meanings, such as desert and cold, and voice recognition software might struggle with words that sound the same, sch as “witch” and “which.”
To give computers a deeper insight into the meanings of words and the relationships between them, researchers and proponents of semantic technology have devised a number of approaches, many of which fall into two broad categories: enhancing the ability of computers to analyze and comprehend language, and making existing content more machine-readable. Examples of the first approach include advanced artificial intelligence and parallel processing technologies designed to give computers the human-style critical thinking skills required to discern between relevant and irrelevant content. The second category includes techniques for labeling content on the Web as well as ontologies — formal descriptions of concepts that may be unique to a specialized domain, such as biology or engineering.
The World Wide Web is a focal point for semantic technology, and many hope to see the emergence of a next generation Web in which knowledge in different forms can more easily be manipulated, discovered, and shared by software agents. This semantic Web, as it has come to be known, was envisioned by the forces behind the original Web as far back as the late 1990s. Though the full potential of the semantic Web has yet to be realized, aspects of semantic technology are already commonplace online. Many search engines, for example, now examine Web pages for special types of metadata, a type of information that describes other information. One type of metadata can specify to a search engine that a series of numbers is a phone number or physical address, while another type might mark a block of text as a user review of a business or product.
Semantic technology could also benefit a large number of industries and academic disciplines. Online advertisers are looking to something called semantic targeting to analyze the content of a Web page and deliver ads relevant to that content. Large corporations and enterprises are eager to eliminate compatibility problems between different information technology systems with software and database architectures that better understand the meaning and context of different content. For academics and researchers, ontologies specific to certain disciplines could allow computers to find and group relevant research on very specialized topics, such as a particular protein marker, allowing humans to spend more time analyzing and conducting research rather than looking for it.