Semantic search

Semantic search seeks to improve search accuracy by understanding searcher intent and the contextual meaning of terms as they appear in the searchable dataspace, whether on the Web or within a closed system, to generate more relevant results. Author Seth Grimes lists "11 approaches that join semantics to search", and Hildebrand et al. provide an overview that lists semantic search systems and identifies other uses of semantics in the search process. Semantic Search systems consider various points including context of search, location, intent, variation of words, synonyms, generalized and specialized queries, concept matching and natural language queries to provide relevant search results. Major web search engines like Google and Bing incorporate some elements of Semantic Search.

Guha et al. distinguish two major forms of search: Navigational and Research. In navigational search, the user is using the search engine as a navigation tool to navigate to a particular intended document. Semantic Search is not applicable to navigational searches. In Research Search, the user provides the search engine with a phrase which is intended to denote an object about which the user is trying to gather/research information. There is no particular document which the user knows about that s/he is trying to get to. Rather, the user is trying to locate a number of documents which together will give him/her the information s/he is trying to find. Semantic Search lends itself well here.

Rather than using ranking algorithms such as Google's PageRank to predict relevancy, Semantic Search uses semantics, or the science of meaning in language, to produce highly relevant search results. In most cases, the goal is to deliver the information queried by a user rather than have a user sort through a list of loosely related keyword results.

Other authors primarily regard semantic search as a set of techniques for retrieving knowledge from richly structured data sources like ontologies as found on the Semantic Web. Such technologies enable the formal articulation of domain knowledge at a high level of expressiveness and could enable the user to specify his intent in more detail at query time.

Disambiguation

In order to understand what a user is searching for, word sense disambiguation must occur. When a term is ambiguous, meaning it can have several meanings (for example, if one considers the lemma "bark", which can be understood as "the sound of a dog," "the skin of a tree," or "a three-masted sailing ship"), the disambiguation process is started, thanks to which the most probable meaning is chosen from all those possible.

Such processes make use of other information present in a semantic analysis system and takes into account the meanings of other words present in the sentence and in the rest of the text. The determination of every meaning, in substance, influences the disambiguation of the others, until a situation of maximum plausibility and coherence is reached for the sentence. All the fundamental information for the disambiguation process, that is, all the knowledge used by the system, is represented in the form of a semantic network, organized on a conceptual basis.

In a structure of this type, every lexical concept coincides therefore with a semantic network node and is linked to others by specific semantic relationships in a hierarchical and hereditary structure. In this way, each concept is enriched with the characteristics and meaning of the nearby nodes.

Every node of the network (called Synset) groups a set of synonyms which represent the same lexical concept (called Synsets) and can contain:

• single lemmata ('seat', 'vacation'; 'work', 'quick'; 'quickly', 'more', etc.)
• compounds ('non-stop', 'abat-jour', 'policeman')
• collocations ('credit card', 'university degree', 'treasury stock', 'go forward', etc.).

The semantic relationships (links), which identify the semantic relationships between the synsets, are the order principals for the organization of the semantic network concepts.

Commonly used searching methodologies

Mäkelä describes five mainly used methodologies:

• RDF Path Traversal - traversing the net formed by the RDF data format.
• Keyword to Concept Mapping
• Graph Patterns - used to formulate patterns for locating interesting connecting paths between resources. Also commonly used in data visualization.
• Logics - by using inference based on OWL
• Fuzzy Concepts, Fuzzy Relations, Fuzzy Logics

Semantic search portals

• GoPubMed - first semantic search engine on the internet - launched in 2002
• Research Gate - The professional network for scientists
• Google
• Hakia
• iGlue - semantic search engine with realtime annotator plugin/bookmarklet which adds a smart layer to every website
• Kosmix – social media semantic search
• Lexxe – (beta in early 2011)
• Swoogle
• Yummly - food & recipe semantic search
• Bing

Enterprise semantic search engines

• Endeca
• Exalead
• Invention Machine
• Sinequa
• salsaDev
• Smartlogic
• Sophia
• OpenText
• Inbenta

See also

• List of Semantic Search Engines
• Semantic web
• Semantic Unification
• Resource Description Framework
• Natural language search engine

References

Several scientific events cover the topic of semantic search explicitly, such as the Semantic Search 2008 Workshop at ESWC'08 and the Workshop on Exploiting Semantic Annotations in Information Retrieval at ECIR'08.