Video Presentation 2016 June

Category: Education

Presentation Description

No description available.


Presentation Transcript

Enhancing Educational Big Data Analytics using Linked Data :

Enhancing Educational Big Data Analytics using Linked Data By : Mohamed M. Maher Supervisor: liliana patricia santacruz

Education Big data Resources:

Education Big data Resources Analytics Using The Data Mining Techniques Classification Patterns discovery Clustering Regression …….

Learning Linked Data Analytics:

Learning Linked Data Analytics Open Data Analytics Using The Data SPARQL Ontology Reasoners Others …….

Research Contribution:

Research Contribution To build a Model platform that will get benefits from the linked data knowledge repositories and query power and the Big Data Mining techniques to do the following : 1. Enhance Big Educational Data analytics processes using the linked data relations Ex. Classifier features selection will depend on the concept ( properties) relations in the linked data ontology 2. Convert the output of educational data descriptive analytics into Linked data ( concepts ,elations) to support the future needs ( Course Recommendation , Students Clustering ,…) Ex. ( the output of the association rules to be converted to semantic relations in the educational ontology )

Pattern Discovery to Ontology Rules Example 1:

Pattern Discovery to Ontology Rules Example 1 80 % of students who went through the chemistry course (CHM230) exam did not answer question No. 3 properly Question No. 3 is assessing the Learning object (LO756) The same situation happened for the that last year with a failure rate of 78% Using the associations rules mining techniques it has been discovered the students succeeded to answer the (LO763) question has studied previously other learning objects (LO762) & (LO751) (LO762) & (LO751) are included in other Separate Courses So either to include these LOs in the current course to (CHM230) according to teaching plans or enforce prerequisites course before it These inferred rules will be translated into a conditional linked data rule in the Ontology Knowledge Repository

Pattern Discovery to Ontology Rules Example2:

Pattern Discovery to Ontology Rules Example2 E ach student Learning Path could be represented by a sequence of courses Sequence mining will find frequent patterns in these sequences like Students Groups with certain credits who take the same courses sequence Courses often taken together in certain semesters (CS100) (IT237, S265) (S345)

Current Status :

Current Status Item / Topic Status Study the Big data in Education Done Study the different Linked Data application in Education Done Study in the Mapping techniques between Big Data analytics and Linked Data Running Design a Dynamic Extensible Higher Education Analytical Model “Classifier” Running Build the Model and its Visualization Not Yet Apply the model on different data sets Not Yet Model Tuning Not Yet

Slide 8:

Thank You

authorStream Live Help