(C) 2017 SPEET Project

SPEET Secretariat

School of Engineering, UAB

08193, Bellaterra, Spain

(34 )935 812 197

Legal Notice

Sitemap

Project Outputs

 

Intellectual Output #1

Data Mining Tool for Academic Data Exploitation:  Literature review and first architecture proposal

What will you find in the document?

This output take the form of a publication document that aims to reflect the necessary aspects implied into the characterisation of a student profile: pedagogical characteristics, teaching learning attitudes, description of the different situations that may reflect problems regarding a normal progress.

 

Also the characterisation the scenario that concerns the application of educational data mining techniques. The data that a student generates while progressing on his/her studies will be synthesised and related to potential profile features.

As per the SPEET project concern, the definition of an IT architecture that is aimed at dealing with such student profile characterisation is also outlined.

This will be of your interest if .....

... you are into an academic management position, as will find out possibilities that data exploration can provide you in relation with the actual academic data you store from your students. 

... you are into data analytics, the academic world as a fantastic scenario where the techniques can be applied with the focus student performance improvement and the early detection of the important drop-out problem.

... you are into the teaching activity you will realise how important may be the information associated  to each student that is being recorded and to realise it is possible to learn t what extent the past student records of the students can be used to predict expected performance on actual subjects.

 

Intellectual Output #2

Data Mining Tool for Academic Data Exploitation:  

Selection of most suitable algorithms

PortadaIO2.jpg

What will you find in the document?

This output take the form of a publication document that aims to reflect the results obtained at SPEET project under the development of the data mining tools are presented. More specifically, two mechanisms have been developed: a clustering/classification scheme of students in terms of academic performance and a drop-out prediction system.

The students’ clustering and classification schemes are presented in detail. More specifically, a description of the considered machine learning algorithms can be found.  Results show how groups of clusters can be automatically identified and how new students can be classified into existing groups with a high accuracy. Finally, the implemented drop-out prediction system is considered by

presenting several algorithms alternatives. In this case, the evaluation of the dropout mechanism is focused on one institution, showing a prediction accuracy around 91 %.

Algorithms presented at this document are available at repositories or inline code format, as accordingly indicated.

This will be of your interest if .....

... you are into an academic management position, as will find out possibilities that data exploration can provide you in relation with the actual academic data you store from your students. 

... you are into data analytics, the academic world as a fantastic scenario where the techniques can be applied with the focus student performance improvement and the early detection of the important drop-out problem.

... you are into the teaching activity you will realise how important may be the information associated  to each student that is being recorded and to realise it is possible to learn t what extent the past student records of the students can be used to predict expected performance on actual subjects.

 

Intellectual Output #3

Data Mining Tool for Academic Data Exploitation:  

Graphical Data analysis and Visualization

Portada_IO3.jpg

What will you find in the document?

This output take the form of a publication document that aims to reflect the results obtained at SPEET project under the development of the data mining tools are presented. More specifically, two mechanisms have been developed: a clustering/classification scheme of students in terms of academic performance and a drop-out prediction system.

The students’ clustering and classification schemes are presented in detail. More specifically, a description of the considered machine learning algorithms can be found.  Results show how groups of clusters can be automatically identified and how new students can be classified into existing groups with a high accuracy. Finally, the implemented drop-out prediction system is considered by

presenting several algorithms alternatives. In this case, the evaluation of the dropout mechanism is focused on one institution, showing a prediction accuracy around 91 %.

Algorithms presented at this document are available at repositories or inline code format, as accordingly indicated.

This will be of your interest if .....

... you are into an academic management position, as will find out possibilities that data exploration can provide you in relation with the actual academic data you store from your students. 

... you are into data analytics, the academic world as a fantastic scenario where the techniques can be applied with the focus student performance improvement and the early detection of the important drop-out problem.

... you are into the teaching activity you will realise how important may be the information associated  to each student that is being recorded and to realise it is possible to learn t what extent the past student records of the students can be used to predict expected performance on actual subjects.

 

Intellectual Output #4

Data Mining Tool for Academic Data Exploitation:  

Publication Report on Engineering Students Profiles

Portada_IO4.jpg

What will you find in the document?

This output take the form of a publication document that aims to reflect the results obtained at SPEET project by means of the application of the deployed data mining tools. More specifically, the application of the student performance tools on the engineering degrees by the participating institutions. The analysis can be used twofold:

  • Example on the use of the different performance analysis tools developed on the project and accessible at the project web tool.

  • Example on the interpretation of the clustering and analysis tools in order to delimit a student profile. 

 

The provided analysis also extends the interpretations aiming at showing similarities country wise, so, finally we can conclude that the tools developed in this project can offer some significant information in detecting different profiles and the relationship between these profiles and categorical variables such as age, admission score, sex, previous studies. 

This will be of your interest if .....

... you are into an academic management position, as will find out possibilities that data exploration can provide you in relation with the actual academic data you store from your students. 

... you are into data analytics, the academic world as a fantastic scenario where the techniques can be applied with the focus student performance improvement and the early detection of the important drop-out problem.

... you are into the teaching activity you will realise how important may be the information associated  to each student that is being recorded and to realise it is possible to learn t what extent the past student records of the students can be used to predict expected performance on actual subjects.

Intellectual Output #5

Data Mining Tool for Academic Data Exploitation:  

Webtool Description and usage

Portada_IO5.jpg

What will you find in the document?

The report corresponding to this output is related to the Data Mining Tools for Academic Data Exploitation, and collects the overall architecture of the SPEET generated WEB-tool, the user manual for the use of WEB-tool and the overview of the output to establish clustering, dropout, and their dependency on stu- dents’ characteristics.

The ultimate goal of SPEET project is the development of an WEB-based tool to disseminate the main intellectual output in form of user-friendly and easily accessible software tool. The WEB-tool is accessible from speet.uab.cat is intended to make accessible by other faculties and schools outside of the SPEET consortium the possibility to make data analysis on students based on the proprietary data after these are organized accordingly.

The WEB-tool developed within SPEET project represent a product that will remain after the end of the project to let European Universities access- ing and comparing their performance metrics. Furthermore, the WEB-tool is a free-of-charge tool that contains the analyses of the SPEET institutions and another engineering institution can autonomously compare with the pre-existing ones to extract comparative analyses, or simply to investigate anomalous dropout situations.

This will be of your interest if .....

... you are into an academic management position, as will find out possibilities that data exploration can provide you in relation with the actual academic data you store from your students. 

... you are into data analytics, the academic world as a fantastic scenario where the techniques can be applied with the focus student performance improvement and the early detection of the important drop-out problem.

... you are into the teaching activity you will realise how important may be the information associated  to each student that is being recorded and to realise it is possible to learn t what extent the past student records of the students can be used to predict expected performance on actual subjects.