langreiter.com plain, simple: CIRCA

LANGREITER.COM plain, simple

START

INDEX

CIRCA

CREATED BY chris • LAST EDITED BY chris 6733 days AGO

Conceptual Information Retrieval and Communication Architecture

"At the heart of the strategy behind CIRCA rests the idea of finding connections between things that are related to one another. This can apply to any type of information object – as long as it is possible to connect the object to a specific concept or idea. [...] In order to determine the associations between objects, a semantic space is defined, within which conceptual distances, or the closeness of association between things may be measured. [...]"

"The Applied Semantics Ontology [tokens/terms/meanings] is the backbone of CIRCA functionality. This massive database stores the human-generated and statistical information that enables the interpretation of text. [...] Each relationship is associated with a strength indicating how close the relationship is. [...] [T]erm analysis and meaning disambiguations are done on the basis of human-estimated probabilities and conceptual relationships provided in the Ontology. [...]"

Relationship types (quoted from the whitepaper as well):

Synonymy/antonymy ("good"/"bad")

Similarity ("gluttonous" is similar to "greedy")

Hypernymy (is a kind of / has kind) ("horse" has kind "Arabian")

Membership ("commissioner" is a member of "commission")

Metonymy (whole/part relations) ("motor vehicle" has part "clutch pedal")

Substance (e.g. "lumber" has substance "wood")

Product (e.g. "Microsoft Corporation" produces "Microsoft Access")

Attribute ("past", "preceding” are attributes of "timing")

Causation (e.g. "travel" causes "displacement" or "motion")

Entailment (e.g. "buying" entails "paying")

Lateral bonds (concepts closely related to one another, but not in one of the other relationships, e.g. "dog" and "dog collar")

"[M]echanisms for automatically generating new relationships from those represented in the basic Ontology have been implemented. [...] The architecture as presented here is related to a well-known semantic network called WordNet [plus probabilistic data and 'lateral relationships that help to organize concepts into coherent groups' used for disambiguation] [...]"

[create Concept Similarity Measurement]

http://www.appliedsemantics.com/pdf/CIRCA_whitepaper.pdf

Note: The exceptionally detailed whitepaper has been removed from the Applied Semantics site after the company was acquired by Google — google-watch.org ~~has~~ had a mirror (now 404).