- Introduction back to ToC
- Scientific Discourse Ontology Description back to ToC
- Cross reference for Scientific Discourse Ontology classes and properties This section provides details for each class and property defined by Scientific Discourse Ontology.
- is equivalent to
- model c
- has super-classes
- :ScientificDiscourse c
- is equivalent to
- sro:Motivation
- has super-classes
- ScientificDiscoursec
- is equivalent to
- sro:Background
- has super-classes
- ScientificDiscoursec
- has sub-classes
- CommonGround c, RelatedWork c
- has super-classes
- Backgroundc
- is equivalent to
- sro:Conclusion
- has super-classes
- ScientificDiscoursec
- has sub-classes
- Limitationsc
- is equivalent to
- sro:Contribution
- has super-classes
- ScientificDiscourse c
- is equivalent to
- sro:Discussion
- has super-classes
- ScientificDiscourse c
- has super-classes
- ScientificDiscoursec
- is equivalent to
- deo:FutureWork
- has super-classes
- ScientificDiscourse c
- has super-classes
- ScientificDiscourse c
- has super-classes
- ScientificDiscourse c
- has super-classes
- Conclusion c
- is equivalent to
- deo:Methods
- has super-classes
- ScientificDiscourse c
- is equivalent to
- deo:ProblemStatement
- has super-classes
- ScientificDiscourse c
- has super-classes
- Background c
- deo:RelatedWork
- is equivalent to
- deo:Results
- has super-classes
- ScientificDiscourse c
- References back to ToC
- [DEO] Discourse Elements Ontology, http://purl.org/spar/deo
- [Liakata 2008] Liakata, M., & Soldatova, L. (2008). Guidelines for the annotation of general scientific concepts. Aberystwyth University, JISC Project Report http://ie-repository.jisc.ac.uk/88
- [Liakata 2010] Liakata, Maria, Simone Teufel, Advaith Siddharthan, y Colin R. Batchelor. Corpora for the Conceptualisation and Zoning of Scientific Papers. In Proceedings of the International Conference on Language Resources and Evaluation, LREC 2010, 17-23 May 2010, Valletta, Malta.http://www.lrec-conf.org/proceedings/lrec2010/summaries/644.html
- [Teufel 2010] The Structure of Scientific Articles: Applications to Citation Indexing and Summarization Simone Teufel (University of Cambridge)
- Acknowledgements back to ToC
The sci-Doc ontology aims to model the rhetorical elements of the scientific discourse. The scientific discourse elements of Sci-Doc are based on the corpora proposed by [Liakata 2008] and extended with our analysis from a set of scientific documents. We have selected the next concepts from the Liakata model: hypothesis, motivation, background, goal, experiment, method, result, conclusion.
Although these concepts cover the vast majority of the scientific discourse elements we added some other concepts that represent the main characteristics of the scientif discourse. These concepts are: model, observation, common ground, contribution, discussion, future work, limitations, problem statement and related work.
For more information about the classes and properties of the ontology check the following sections.
As we stated before, the Sci-Doc ontology aims to model the rhetorical elements of the scientific discourse. In this sense, we follow the corpora proposed by [Liakata 2010] in order to annotate the scientific discourse. They propose the use of the next concepts for conceptualize the scientific discourse: hypothesis, motivation, background, goal, experiment, method, result and conclusion. We have extended this model with new classes that are not covered by it. These new classes are: common ground, related work and limitations. Next we explain these new classes.
The Sci-Doc ontology imports the Deo ontology [DEO] due to the Deo ontology provides the vast majority of the classes that are needed to model the scientific discourse. For example, we use the Background class to identify what is the essential knowledge for understanding the problem. The class Contribution annotates a description of the part that this publication plays in the overall field. Notice that [DEO] covers the most relevant rhetorical elements, but we should extend it with some specific classes according to the concrete domain. Specifically we need to add the classes related to the Liakata model that are not covered by the Deo ontology, namely: experiment, goal, hypothesis and observation. Figure 1 shows a schema of the ontology classes.
First, we want to distinguish between the common ground (general concepts about the domain) and the related work (other works that cover the same topic). These two concepts are subclasses of the background class. Next, a scientific document has goals and at least one hypothesis. These classes are subsclasses of the main class (ScientificDiscourse). Another characteristic in the scientific discourse is the experiment carried out. We have included a class named Experiment that identify this element. Finally, we consider interesting to point out the limitations of the scientific work, in this sense we have included the class Limitations (is subclass of conclusion) that points out the limitations of the author' work.
Classes
Modelc back to ToC or Class ToC
IRI: http://purl.org/drinventor/sci-doc#Model
Motivationc back to ToC or Class ToC
IRI: http://purl.org/drinventor/sci-doc#Motivation
The stimulus for achieving the goal of the investigation, the reason to carry out the investigation [Liakata 2008]
Backgroundc back to ToC or Class ToC
IRI: http://purl.org/drinventor/sci-doc#Background
Theoterical information about the domain.
According to the annotation scheme proposed by Liakata et. al 2008, generally accepted background knowledge and previous work accepted background knowledge and previous work.
According to [DEO] ontology: Presentation of information that is essential for understanding the situation or problem that is the subject of the publication. In a journal article, the background is usually part of the Introduction, but may be present as separated section.
CommonGroundc back to ToC or Class ToC
IRI: http://purl.org/drinventor/sci-doc#CommonGround
Knowledge that is known by everyone in the domain. According to [Teufel 2010] Common Ground is defined as "no knowledge claim raised".
Conclusionc back to ToC or Class ToC
IRI: http://purl.org/drinventor/sci-doc#Conclusion
The most important take-home message of the study.
According to [Liakata 2008], conclusions are statements inferred from observations and results, which support or reject a research hypothesis or summarise the findings of an investigation.
Contributionc back to ToC or Class ToC
IRI: http://purl.org/drinventor/sci-doc#Contribution
Indicates why the work is important.
The Deo ontology [DEO] defines contribution as a description of the part that this publication plays in the overall field.
Discussionc back to ToC or Class ToC
IRI: http://purl.org/drinventor/sci-doc#Discussion
As stated in Deo [DEO], discussion is an interpretation of the results obtained and an analysis of their significance, in support of conclusions.
Experimentc back to ToC or Class ToC
IRI: http://purl.org/drinventor/sci-doc#Experiment
How the study was structured and how it was carried out (experimental setup, metrics and details)
FutureWorkc back to ToC or Class ToC
IRI: http://purl.org/drinventor/sci-doc#FutureWork
The perspective of the work. Suggestions about how to improve the work (own or general)
Goalc back to ToC or Class ToC
IRI: http://purl.org/drinventor/sci-doc#Goal
According to [Liakata 2008]: A goal of an investigation is the target state of the investigation where intended discoveries are made, approaches are tested, problems are demonstrated, tasks formulated etc.
Hypothesisc back to ToC or Class ToC
IRI: http://purl.org/drinventor/sci-doc#Hypothesis
A hypothesis is an educated guess about how things work. A hypothesis should be something that you can test.
Limitationsc back to ToC or Class ToC
IRI: http://purl.org/drinventor/sci-doc#Limitations
Limitations of the author's work
Methodc back to ToC or Class ToC
IRI: http://purl.org/drinventor/sci-doc#Method
How the author performed the experiment. Technique/s used to achieve the goal/s.
ProblemStatementc back to ToC or Class ToC
IRI: http://purl.org/drinventor/sci-doc#ProblemStatement
States a specific research question.
Deo ontology [DEO] defines this class as a concise description of the issues that needed to be addressed by a work described in the document.
RelatedWorkc back to ToC or Class ToC
IRI: http://purl.org/drinventor/sci-doc#RelatedWork
Work done by others in the same domain, this includes also the related work of the author's own work.
Resultc back to ToC or Class ToC
IRI: http://purl.org/drinventor/sci-doc#Result
Data collected during experimentation.
According to Deo ontology definition [DEO] : The report of the specific findings of an investigation, given without discussion or conclusion being drawn.
The authors would like to thanks Silvio Peroni for developing LODE, a Live OWL Documentation Environment used for representing the Cross Referencing Section of this document and Daniel Garijo for developing the script used to create the template of this document.