The Mentions Ontology (MO) 1.0
Specification of MO Core - 14 March 2016
|
Abstract
Mentions Ontology (MO) is an ontology for annotating the elements mentioned in different texts. It is based in the Modular Unified Taggin Ontology (MUTO) due to its similarities with tagging and for reusability reasons.
Table of Contents
- Introduction
- Mentions Ontology (MO) at a glance
- Examples
- Mentions Ontology (MO) cross-reference
- External Vocabulary References
- Acknowledgments
- References
- Recent Changes
Introduction
The following specification is a formal description of metadata schema proposal that can be applied to any "text based" content. The aim of the following section is to introduce both Semantic Web and Content Management experts to the topic and goals of the ontology, and to provide the basic knowledge to comprehend the technical part of the specification.
An important note is that Mentions Ontology is not a complete model of Mentioning. It extends the concepts of the main Modular Unified Tagging Ontology (MUTO) and defines concepts for adding the mentioning capabilities to a Tagging ontology.
Background [back to top]
The work done with this ontology is based on a project called InnoWEB (at Mondragon Unibertsitatea). Part of this project aimes to find Information and Comunication Technologies (ICTs) that help to improve innovation processes. A first approach to this objective was made identifying the bennefits that Semantic Web can bring to Innovation processes [2].
A case study of this ontology was to find resources mentioned in Ideas (from an Innovation Process platform, more specifically in an Idea Management System) and News Items (from a Technology Watch platform) about the same topics [1]. Using the methodology described in the research, it is possible to find 2 different type of relationships:
- Idea-idea relationships: we can identify the 10 most similar ideas for each idea.
- News items-idea relationships: we can identify the 10 most similar ideas for each news items.
An automatic way was proposed to find those relationships using the DBPedia Spotlight API. The most successfull aproach was found taking mentions into account, both from the title and the body of the content, along with the similarity score provided by DBPedia Spotlight API. Therefore, and in order to represent this approach semantically, Mentions Ontology (MO) was built.
What is Mentions Ontology (MO) for? [back to top]
The goals of Mentions Ontology to achieve as a data schema are:
Extend the notion of MUTO's Tagging to 'mentioning', that can have a Surface Form, Offset and a similarity Score in order to identify where the resource has been mentioned in a text and how probable it is to be that resource.
It is also possible to identify where the resource has been mentioned, for example, whether a resource has been mentioned in the title of an idea or its body.
Mentions Ontology (MO) at a glance [back to top]
An a-z index of Mentions Ontology (MO) terms, by class (categories or types) and by property.
The class diagram of Mentions Ontology (MO).
Classes: | Mention | AutoMention |
Properties: | mentionedIn | offset | score | surfaceForm |
Example [back to top]
The following diagram ilustrates an example of adding mentions to an idea from an imaginary Idea Management System (IMS) at example.org. The idea has a title (“Card-lock for laboratory door.”) and a body (“I often forget keys to the laboratory, so i think it would be great to install a card access lock with a reader for chips in spanish ID”).
According to the diagram, a user has identified that the idea has mentioned 2 times the same resource: “laboratory”. Therefore, the user has added 2 mentions of the same resource ("http://dbpedia.org/resource/Laboratory") with the same surface form ("laboratory"), one to the title of the idea (dcterms:title) with offset 14 and another to the content of the idea (gi2mo:content) with offset 27. He has also labeled the mention as “Laboratory”.
The automatic mentioning system has automatically recognized the resource "http://dbpedia.org/resource/Documento_Nacional_de_Identidad" from the surface form "spanish ID" within the content of the idea, at offset 125. Therefore, it has created a new AutoMention linking it to the resource. According to the system, there is a 95% of probability for the surface form to be the retrieved resource, so it has set the Score to 0.95. He has also identified the label of the resource and has set it to “Documento Nacional de Identidad”. Different resources mentioned in the idea have been identified, both manually and automatically. This can be done with all the ideas of the IMS. A graphic of a simple example of an use case of the Mentions Ontology (MO) The Mentions Ontology (MO) introduces the following classes and properties. Classes: | Mention | AutoMention |
Properties: | mentionedIn | offset | score | surfaceForm |
[#] [back to top] [#] [back to top] [#] [back to top] [#] [back to top] [#] [back to top] [#] [back to top] This ontology has been created to add "mentions" feature to the Modular Unified Tagging Ontology (MUTO). The example of this ontology also uses Gi2MO Ontology for idea annotation. This work has been conducted in the context of the Innoweb project in Mondragon Unibertsitatea. [1] Pending: Semantic Annotations to enhance Innovation and Technology Watch Interoperability. [2] Perez, A., Larrinaga, F., Curry, E. (2013). The Role of Linked Data and Semantic-Technologies for Sustainability Idea Management. In Software Engineering and Formal Methods (pp. 306-312). Springer International Publishing. [BibTex] [3] Lohmann, S., Díaz, P., Aedo, I.: MUTO: The Modular Unified Tagging Ontology. Proceedings of the 7th International Conference on Semantic Systems (I-SEMANTICS 2011), pp. 95-104. ACM, New York, NY, USA, 2011. [BibTex] [4] A. Westerski, Gi2MO: Interoperability, Linking and Filtering in Idea Management Systems, in Extended Semantic Web Conference 2011. PhD Symposium Poster., Heraklion, Greece, 2011. [BibTex] This work is licensed under a Creative Commons Attribution License. This copyright applies to the Mentions Ontology (MO) Specification and accompanying documentation in RDF. Regarding underlying technology, the Mentions Ontology (MO) uses W3C's RDF technology, an open Web standard that can be freely used by anyone.
@prefix mo: <http://w3id.org/mo/core#> .
@prefix gi2mo: <http://purl.org/gi2mo/ns#> .
@prefix dcterms: <http://purl.org/dc/terms/> .
<http://example.com/ideas/idea1> a <http://purl.org/gi2mo/ns#Idea>;
dcterms:title "Card-lock for laboratory door";
gi2mo:content "I often forget keys to the laboratory, so i think it would be great to install a card access lock with a reader for chips in spanish ID";
<http://example.org/tagging/taggin1> a <http://purl.org/muto/core#Tagging>;
muto:taggedResource <http://example.com/ideas/idea1>;
muto:hasTag <http://example.org/tag/tag1>,
<http://example.org/tag/tag2>,
<http://example.org/tag/tag3>;
<http://example.org/tag/tag1> a <http://w3id.org/mo/core#Mention>;
muto:tagLabel "Laboratory";
muto:tagMeaning <http://dbpedia.org/resource/Laboratory>;
mo:offset 14;
mo:surfaceForm "laboratory";
mo:mentionedInProperty <http://purl.org/dc/terms/title>;
muto:nextTag <http://example.org/tag/tag2>.
<http://example.org/tag/tag2> a <http://w3id.org/mo/core#Mention>;
muto:tagLabel "Laboratory";
muto:tagMeaning <http://dbpedia.org/resource/Laboratory>;
mo:offset 27;
mo:surfaceForm "laboratory";
mo:mentionedInProperty <http://purl.org/gi2mo/ns#content>;
muto:nextTag <http://example.org/tag/tag3>.
<http://example.org/tag/tag3> a <http://w3id.org/mo/core#AutoMention>;
muto:tagLabel "Documento Nacional de Identidad";
muto:autoMeaning <http://dbpedia.org/resource/Documento_Nacional_de_Identidad>;
mo:offset 125;
mo:surfaceForm "spanish ID";
mo:mentionedInProperty <http://purl.org/gi2mo/ns#content>;
mo:score 0.95 .
Mentions Ontology (MO) cross-reference: Listing the Mentions Ontology (MO) Classes and Properties [back to top]
Classes and Properties (full detail)
Classes [back to top]
Class: Mention
Mention - Mention is a tag associated to a text, an specific string (or surface form) in that text and to the offset of the surface form in the text.
URI:
http://w3id.org/mo/core#Mention
Properties include:
Offset |
Surface form
Sub class of
muto:Tag
OWL Class
Class: AutoMention
Automatic Mention - An automatic mention is a mention that is automatically associated with a resource (e.g. by a text annotation system like DBPedia Spotlight), i.e. it is not entered by a human being.
URI:
http://w3id.org/mo/core#AutoMention
Properties include:
score
Sub class of
Mention |
muto:AutoTag
OWL Class
Properties
Property: Mentioned In
Mentioned In - The property of the source text where the Mention has been mentioned.
URI:
http://w3id.org/mo/core#mentionedIn
Domain:
Mention
Object Property
Property: Offset
Offset - The position (offset) of the surface form string in the source text.
URI:
http://w3id.org/mo/core#offset
Domain:
Mention
Range:
xsd:integer
Datatype Property
Property: Score
Score - The score the automatic system gave to the surface form to be the giving resource.
URI:
http://w3id.org/mo/core#score
Domain:
Automatic Mention
Range:
xsd:float
Datatype Property
Property: Surface Form
Surface form - The surface form is the string that has been identified to mention a resource in the text.
URI:
http://w3id.org/mo/core#surfaceForm
Domain:
Mention
Range:
xsd:string
Datatype Property
External Vocabulary References [back to top]
Acknowledgments [back to top]
References [back to top]
Recent Changes [back to top]
======================
Version 1.0 (Mar 2016)
======================