Le Ngoc L., Tireau A., Venkatesan A., Neveu P., Larmande Pierre. (2016). Development of a knowledge system for Big Data : case study to plant phenotyping data. In :
WIMS'16 : international conference on web intelligence, mining and semantics. New York : ACM, art. no 27 [9 p.]. (International Conference Proceedings Series). WIMS'16 : International Conference on Web Intelligence, Mining and Semantics, Nîmes (FRA), 2016/06/13-15. ISBN 978-1-4503-4056-4.
Titre du document
Development of a knowledge system for Big Data : case study to plant phenotyping data
Année de publication
2016
Type de document
Partie d'ouvrage
Auteurs
Le Ngoc L., Tireau A., Venkatesan A., Neveu P., Larmande Pierre
In
WIMS'16 : international conference on web intelligence, mining and semantics
Source
New York : ACM, 2016,
art. no 27 [9 p.] (International Conference Proceedings Series). ISBN 978-1-4503-4056-4
Colloque
WIMS'16 : International Conference on Web Intelligence, Mining and Semantics, Nîmes (FRA), 2016/06/13-15
In the recent years, the data deluge in many areas of scientific research brings challenges in the treatment and improvement of agricultural data. Research in bioinformatics field does not outside this trend. This paper presents some approaches aiming to solve the Big Data problem by combining the increase in semantic search capacity on existing data in the plant research laboratories. This helps us to strengthen user experiments on the data obtained in this research by infering new knowledge. To achieve this, there exist several approaches having different characteristics and using different platforms. Nevertheless, we can summarize it in two main directions: the query re-writing and data transformation to RDF graphs. In reality, we can solve the problem from origin of increasing capacity on semantic data with triplets. Thus, data transformation to RDF graphs direction was chosen to work on the practical part. However, the synchronization data in the same format is required before processing the triplets because our current data are heterogeneous. The data obtained for triplets are larger that regular triplestores could manage. So we evaluate some of them thus we can compare the benefits and drawbacks of each and choose the best system for our problem.
Plan de classement
Sciences du monde végétal [076]
;
Informatique [122]
