Semantic Web Platforms for Bioinformatics and Life Sciences

Background

One of the main problems with natural languages is ambiguity: the same word can have different meanings and the same meaning can be expressed using different words. The ambiguity embedded in human languages was of course targeted by many fields throughout history, beginning with philosophy, linguistics and more recently by information science and then computer science.

A way of dealing with this problem, leveraging an information science approach, is through the use of controlled vocabularies, where to a particular concept is assigned an official term in order to avoid any confusion. In these tools simple relationships are defined like synonyms and antonyms. There are variations on the scope of these thesauri: some take a broad approach in order to classify all human knowledge (like the Library of Congress Subject Headings), while others go more in depth in a particular field (like the Medical Subject Headings or MeSH). Others take a semantic network approach and are called metathesauri, like the Unified Medical Language System (UMLS).

In addition, these controlled vocabularies define a hierarchy of terms going from the most general, broader term, to the more specific narrower term. The data structure in use is usually a tree, where the broader term lies on top and the narrower term on the bottom.

Ontologies are more complex than controlled vocabularies especially in the complexity of their relationships and in their architecture, something that allows algorithms to infer knowledge that is not explicitly defined in the ontology itself.

Unlike those alternative hierarchical views of concepts such as taxonomies, ontologies often have a graph structure characterized by complex relationships.