Semantic annotation
0 likes287 views
The document outlines the development of a semantic annotation framework aimed at enhancing data understanding through the use of ontologies and RDF technology. It details the project structure, theoretical and practical solutions, tools utilized, achievements made, and future aspirations to integrate machine learning for improved intelligence. The project, titled Agrold, is designed to integrate data from various plant-centric sources to facilitate research in agriculture.
![Introduction
Project
Theoretical
Solution
Practical
solution
Achievements
Demo
Conclusion
Projet
AgroLD projet[1] is a RDF knowledge base designed to
integrate data from various publically important plant centric
data sources such as Gramene, Oryzabase, TAIR.
In order to improve solutions in this erea of crops, researchers
have to understand data coming from various fields, which are
highly distributed
Figure – Understanding data from other domain
3 octobre 2018 4 / 13](https://image.slidesharecdn.com/presentationtpe-181003213031/85/Semantic-annotation-4-320.jpg)
![Introduction
Project
Theoretical
Solution
Practical
solution
Achievements
Demo
Conclusion
Semantic annotation
Annotation : Annotation is a set of explanatory notes that
accompanies a text[5] .
In information science, an annotation is a graphic or textual
information linked to a document or simply placed on top of it.
Semantic : Using semantic web technologies
URI :Uniform Resource Identifier
(a) Concept matching (b) Semantic annotation
Figure – Annotation system
3 octobre 2018 5 / 13](https://image.slidesharecdn.com/presentationtpe-181003213031/85/Semantic-annotation-5-320.jpg)
![Introduction
Project
Theoretical
Solution
Practical
solution
Achievements
Demo
Conclusion
Tools
To achieve this work, many tools were used :
Programming language : Python ;
Python Framework : Flask[4]
CSS Framework : Bootstrap
NLP : Python NLTK[2]
Query Ontology : Python RDFLib
Data storage : Elasticsearch[3]
Versioning system : Git + Github
3 octobre 2018 9 / 13](https://image.slidesharecdn.com/presentationtpe-181003213031/85/Semantic-annotation-9-320.jpg)
![Annexe
For Further Reading
Références
[1] Agrold :home.
http://volvestre.cirad.fr:8080/agrold/. (Accessed
on 10/03/2018).
[2] Natural language toolkit — nltk 3.3 documentation.
https://www.nltk.org/. (Accessed on 10/03/2018).
[3] Open source search & analytics · elasticsearch | elastic.
https://www.elastic.co/. (Accessed on 10/03/2018).
[4] Welcome | flask (a python microframework).
http://flask.pocoo.org/. (Accessed on 10/03/2018).
[5] Yongxin Liao, Mario Lezoche, Hervé Panetto, and Nacer
Boudjlida. Why, where and how to use semantic annotation
for systems interoperability. In 1st UNITE Doctoral
Symposium, pages 71–78, 2011.
3 octobre 2018 13 / 13](https://image.slidesharecdn.com/presentationtpe-181003213031/85/Semantic-annotation-13-320.jpg)
Semantic annotation
- 1. Introduction Project Theoretical Solution Practical solution Achievements Demo Conclusion TRAVAUX PRATIQUE ENCADRES Theme : Develop a semantic annotation framework to annotate semantically some datasets using several ontologies Présenté par : SONFACK SOUNCHIO Serge Encadrant : Dr Pierre LARMANDE 3 octobre 2018 1 / 13
- 2. Introduction Project Theoretical Solution Practical solution Achievements Demo Conclusion 1 Introduction 2 Project 3 Theoretical Solution 4 Practical solution 5 Achievements 6 Demo 7 Conclusion 3 octobre 2018 2 / 13
- 3. Introduction Project Theoretical Solution Practical solution Achievements Demo Conclusion Introduction Internet, the Web and semantic Web has brought us in the world of data and sharing, collaborating are important factors of our growth today. One of the key issues related to this fact is the understanding of these data. Figure – A world of data and collaboration 3 octobre 2018 3 / 13
- 4. Introduction Project Theoretical Solution Practical solution Achievements Demo Conclusion Projet AgroLD projet[1] is a RDF knowledge base designed to integrate data from various publically important plant centric data sources such as Gramene, Oryzabase, TAIR. In order to improve solutions in this erea of crops, researchers have to understand data coming from various fields, which are highly distributed Figure – Understanding data from other domain 3 octobre 2018 4 / 13
- 5. Introduction Project Theoretical Solution Practical solution Achievements Demo Conclusion Semantic annotation Annotation : Annotation is a set of explanatory notes that accompanies a text[5] . In information science, an annotation is a graphic or textual information linked to a document or simply placed on top of it. Semantic : Using semantic web technologies URI :Uniform Resource Identifier (a) Concept matching (b) Semantic annotation Figure – Annotation system 3 octobre 2018 5 / 13
- 6. Introduction Project Theoretical Solution Practical solution Achievements Demo Conclusion Concepts Ontology :is a formal, explicit specification of a shared conceptualisation. Knowledge base system RDF : Resource Description Framework Figure – Architecture 3 octobre 2018 6 / 13
- 7. Introduction Project Theoretical Solution Practical solution Achievements Demo Conclusion Architecture GUI : User interface - API : Extending software Query processing - Storage and Data access Document management Figure – Architecture 3 octobre 2018 7 / 13
- 8. Introduction Project Theoretical Solution Practical solution Achievements Demo Conclusion How it works Storing Ontology = RDFLib RDF extration (from URL or owl file) to JSON format Elasticsearch Indexing - Type - Documents Query processing regular expression stopword removal tokenize stemming Access Data (access) constructed query Elasticsearch search API Access Data ( process result) using string similarity algorithm to classify the result API Python Flask REST API 3 octobre 2018 8 / 13
- 9. Introduction Project Theoretical Solution Practical solution Achievements Demo Conclusion Tools To achieve this work, many tools were used : Programming language : Python ; Python Framework : Flask[4] CSS Framework : Bootstrap NLP : Python NLTK[2] Query Ontology : Python RDFLib Data storage : Elasticsearch[3] Versioning system : Git + Github 3 octobre 2018 9 / 13
- 10. Introduction Project Theoretical Solution Practical solution Achievements Demo Conclusion Achievements Storing Ontologies in document base management system Concepts matching In local DMS Online Search concepts using ID Semantic annotation API To be done : Storing annotation ( In one or many graph ) Figure – Work done 3 octobre 2018 10 / 13
- 11. Introduction Project Theoretical Solution Practical solution Achievements Demo Conclusion (a) Home page presenting API (b) User interface Figure – Some views 3 octobre 2018 11 / 13
- 12. Introduction Project Theoretical Solution Practical solution Achievements Demo Conclusion Reminder We were asked to develop a software that will be used to annotate using ontologies. Work done We develop a software for semantic annotation using Python RDFLib library, web technologies and Elasticsearch for indexing data extracted from ontologies. Future See how to integrate Machine Learning in other to make the software intelligent and more accurate 3 octobre 2018 12 / 13
- 13. Annexe For Further Reading Références [1] Agrold :home. http://volvestre.cirad.fr:8080/agrold/. (Accessed on 10/03/2018). [2] Natural language toolkit — nltk 3.3 documentation. https://www.nltk.org/. (Accessed on 10/03/2018). [3] Open source search & analytics · elasticsearch | elastic. https://www.elastic.co/. (Accessed on 10/03/2018). [4] Welcome | flask (a python microframework). http://flask.pocoo.org/. (Accessed on 10/03/2018). [5] Yongxin Liao, Mario Lezoche, Hervé Panetto, and Nacer Boudjlida. Why, where and how to use semantic annotation for systems interoperability. In 1st UNITE Doctoral Symposium, pages 71–78, 2011. 3 octobre 2018 13 / 13