synonym

[zhengj2007]

From @zhengj2007 on July 31, 2015 21:39

From dosu...@gmail.com on March 10, 2011 09:53:44

label for the new term

synonym

Background

This is intended as an IAO mapping for the OBO tag synonym.

This tag is used to relate ontology terms to the language that people use. But that language is not always precise. Words do not come with necessary and sufficient conditions to specify their referents. In regular language, and at least some scientific language, there is no way to define crisp boundaries for the referent of a word that will satisfy most users conception of what that word refers to.

Even within a single scientific community there are often subtle distinctions of usage. For a term used in anatomy, biologists often differ in how they define spatial and or temporal boundaries for the referents of a term. Or they might differ slightly in what subclasses they consider should be subsumed by a class. Or they may use variants of a term to refer to the same thing, or some completely different term for identically or similarly defined structures . We can't possibly add terms for all such cases and deal with the resulting overlapping partonomies or fantastically detailed classifications. Such an ontology would be unmaintainable and probably unusable.

So, for a whole set of closely related terms we choose one definition and and one term and add the rest as synonyms. That way, users can find candidate terms by searching with a term that is familiar to them. It may be useful to further classify (scope) synonyms as having spatial, temporal boundaries or class referents that are broader, narrower, the same as or just slightly different from the referents of the ontology term. But note that only where we can link to a particular usage can we be precise about this (in reference X Michael Ashburner uses term Y as an exact synonym of term Z). The rest of the time, scoping synonyms in this way can only be a matter of judgment about usage by the community of users. Given that, any definitions using quantifiers will be useless. Instead, we need qualitative terms like 'generally' or 'largely'.

Finally, a note on definitions

Defining these synonyms as ontology terms is far from trivial because it implicitly requires some stance on what words in regular language actually refer to. This is not exactly an uncontroversial subject. I have neither the time nor inclination to get involved in a long debate on this matter before we can add this term to the IAO

Textual definition

def: "Information artifact that links an ontology term to a term in language that, within some specified discipline or publication, has referents that largely or completely correspond to the referents of the ontology term." comment: "Publication can be specified directly via a database cross reference, domain Please add an example of usage for that term Please see several 10s of thousands off instances of use in OBO ontologies.

Original issue: http://code.google.com/p/information-artifact-ontology/issues/detail?id=103

Copied from original issue: information-artifact-ontology/IAO#103

[zhengj2007]

From dosu...@gmail.com on March 10, 2011 06:54:31

Note - I have not added this to the existing ticket - https://code.google.com/p/information-artifact-ontology/issues/detail?id=94 , becuase I don't believe that synonyms are alternate labels.

[zhengj2007]

From dosu...@gmail.com on May 05, 2011 08:54:34

For discussion of relationship to 'alternative term' see: https://code.google.com/p/information-artifact-ontology/issues/detail?id=106#c1

[zhengj2007]

From mcour...@gmail.com on May 05, 2011 14:15:51

We intended alternative terms to be exact synonyms, or exact synonyms per community of usage.

Examples of the former:

• hemagglutination inhibition assay has alternative term HAI (we often us alternative terms properties for acronyms, such as PCR etc)
• DMS structure mapping assay has alternative term single nucleotide resolution mapping assay using DMS probe

Examples of the latter:

• we had the case of "probe" being used as "synonyms" to denote either reporter or detector in 2 different microarray communities. We would like to add probe as alternative term on detector and reporter and add a tag as to in which community this is true.
• serum neutralization of viral infectivity assay has alternative term "influenza neutralization assay", with an editor note "MC, 20100217: I added influenza neutralization assay as alternative term. This should be tagged with a community specific value, eg "influenza ontology" or else."

As you can see from those, we have the same intent, provide users with terms they are familiar with, even if specific of a community. If there is any confusion about the IAO alternative term, we may need to update the definition, but I think the intent is the same than what you are trying to achieve with synonym.

[zhengj2007]

From dosu...@gmail.com on May 11, 2011 01:42:36

I think your examples provide a good case for IAO adopting scoped synonyms!

"serum neutralization of viral infectivity assay has alternative term "influenza neutralization assay". Given how many types of viruses might be neutralized by serum, it sounds like one is a subclass of the other. And if not - that one should be a broad synonym of the other.

[zhengj2007]

From dosu...@gmail.com on May 11, 2011 02:29:35

I’d argue that these annotation properties have two possible functions: a. To provide a bridge to language – in all its messiness. This is essential for the ontology to be easily searchable. b. To provide alternative labels for classes/properties – suitable as an aid to display / querying in specific contexts.

I think that ‘synonym’ works well for case ‘a’, and that dictionary definitions of synonym (see below) are consistent with the use of scoping (see clauses ‘or very similar to’ or ‘means exactly or nearly the same’). Based on dictionary definitions / connotations for the word ‘term’, ‘alternative term’ might be better label for case b. But note that some cases may have both functions. So, I think the best way to deal with this is with this is to use annotate ‘synonym’ with some standard annotation property indicating suitability for use as an alternative term.

Dictionary definitions: synonym (wiktionary): A word or phrase with a meaning that is the same as or very similar to, another word or phrase synonym (OED): 1. A word or phrase that means exactly or nearly the same as another in the same language (e.g. shut and close). 2. A word denoting the same thing as another but suitable for a different context (e.g. serpent for snake). 3. A word equivalent to another in some but not all senses (e.g. ship and vessel). term (Wiktionary): A word or phrase, especially one from a specialized area of knowledge. term (OED): A word used to express a definite concept, esp, in a particular branch of study etc. (a technical term). terminology (Wiktionary): A system of specialized terms. terminology (OED): 1. The system of terms used in a particular subject. 2. The science of the proper use of terms.

[zhengj2007]

From dosu...@gmail.com on July 29, 2011 13:19:51

1. Label for IAO mapping of 'OBO synonym'

It is clear that synonym is not an acceptable label for some people involved. Even if these are retained as OBO community labels, the rdfs label still needs to be acceptable chosen still needs to be acceptable to the OBO community.

We need to find a label that is usable: suggestions 'near synonym' (?) 'closely related term' ... ?

1. Definitions.

The definition first suggested in this ticket is not suitable. Here's why:

Two researchers define the boundary of the cerebellum slightly differently. If we were to define ontology classes for these then it is true that

(a) the extensions of these classes are disjoint (b) most people would say that the two classes have nearly the same meaning

I think they would do so because two instances, one from each class, would be very similar to each other.

Because of lack of agreement on a label, we will use the label foo:

name: foo def: "A word or phrase that, within some paper or community, has exactly or nearly the same meaning as this ontology term." comment: It is important when recording foo, to record the paper or community where this is the case. more formal def: "A word or phrase that, within some paper or community that has exactly or nearly the same set of referents as this ontology term, or has a set of referents whose members are very similar to the referents of this ontology term."

name: exact foo def: "A word or phrase that, within some paper or community, has exactly the same meaning as this ontology term." comment: It is important when recording exact foo, to record the paper or community where this is the case. SubPropertyOf: foo

name: broad foo def: “A word or phrase that, within some paper or community, has a slightly broader meaning than this ontology term.” comment: It is important when recording broad foo, to record the paper or community where this is the case. SubPropertyOf: foo

name: narrow foo def: “A word or phrase that, within some paper or community, has a slightly narrower meaning than this ontology term.” comment: It is important when narrow broad foo, to record the paper or community where this is the case. SubPropertyOf: foo

name: related foo def: “A word or phrase that, within some paper or community, has nearly the same meaning as this ontology term, but where the meaning is not simply broader or narrower.” comment: It is important when recording related foo, to record the paper or community where this is the case. SubPropertyOf: foo

[nicolevasilevsky]

@zhengj2007 I think this has been covered on our OBO operations call today and will be addressed by PR https://github.com/information-artifact-ontology/ontology-metadata/pull/66

I think this can be closed?

[zhengj2007]

@nicolevasilevsky OK to close.