TLAD (Teaching, Learning and Assessment of Databases)
Friday, 15 Jul 2011, 16:47
Visible to anyone in the world
Edited by Christopher Douce, Tuesday, 15 Nov 2011, 11:36
I always enjoy visiting Manchester. Having spent many years there (both as an undergraduate and a postgraduate) visiting Manchester almost feels as if I'm coming home. Plus, travelling directly to the computer science building (as a computer scientist) feels as if I'm returning to a 'spiritual home'!
The reason for my most recent visit was to attend the 9th TLAD workshop, which was all about the teaching, learning and assessment of databases. Below is a summary of the event. I hope it is useful for someone.
Paper Session 1 - Data Mining
The first paper that was presented during the day was entitled 'Teaching Oracle Data Miner using Virtual Machines', by Qicheng Yu and Preeti Patel, both from London Metropolitan University, and expertly presented by Preeti.
Preeti made the point that employers are demanding up to date practical and technical skills. The issue of employability was, of course, an issue that was discussed within an earlier HEA event entitled Enhancing the Employability of Computing Students. To address the important issue of technical skills, educators are necessarily faced with the challenge of how to enable learners to make use of industrial tools and products. One of the solutions (in the area of database and data mining education) was to make use of virtual machine technology, such as the Microsoft Virtual PC, Oracle Virtualbox and a product called VMWare Player.
Some of the challenges that have to be addressed are technical i.e. how to share drives through a host computer, and financial or legal challenges, such as how to make sure that any solutions are correctly licenced. Preeti pointed us to a product known as WEKA, which I had never heard of, but it seemed that many of the audience had!
The second paper of the morning was by Hongbo Du from the University of Buckingham. His paper entitled, 'Data mining project: a critical element in teaching learning and assessment of a data mining module' won the 'best paper' prize of the workshop. Hongbo mentioned what seemed to be an important paper in the area, namely, something called the CRISP-DM standard (wikipedia), which enable students to gain an understanding of the data mining process. Hongbo also presented a general framework for assessment (within his paper) which drew upon this methodology before presenting three different case studies.
The key points that I took away from his presentation was: group work is important (but very often students don't want to do this since they might want to own their own scores completely), and that the domain of application is very important too, and represents a substantial area of complexity that students need to necessarily grapple with.
The final presentation of the first session, entitled, 'Making data warehousing accessible' was by Tony Valsamidis from the University of Greenwich. Tony began by presenting some of the problems, such as the availability of large data sets (an issue that I shall return to later), unwieldy query languages and unfamiliar domain and data models.
Tony spoke of a number of different technologies and techniques, some I had used in anger (such as Visual Basic for Applications), others I had heard of but had forgotten the meaning of (such as OLAP). An important issue that was raised was that of data sanitization: when you move data between different systems, things might not be directly compatible, so database practitioners might have to design some magic data transformations.
Paper Session 2 - Database and the Cloud
The fourth presentation was rather different. Mark Dorling from Langley Grammar School described his teaching practice and association with a project called Digital School House.
Mark trained as a primary school teacher but he is now working within the secondary sector. He described how he helps students to understand the key concepts of data, information and logical operators, sometimes applying kinaesthetic learning techniques (i.e. movement). Mark also described the use of something called independent learning videos (allowing students to remember how to use elements of applications). This reminded me of the industrial term of 'on-demand e-learning', where learners can call up screen casts or bite sized presentations about how to carry out or complete different tasks.
Mark also showed how secondary school pupils could make use of cloud application, such as Google Spreadsheet, to enable an entire class to enter data (and for the data to magically appear on an interactive whiteboard). Mark's presentation (and paper) got me thinking about how I might potentially adopt some of the pedagogic techniques that he described into my own teaching practices.
Mark's description of how he makes use of Google Spreadsheet lead us directly to Clare Stanier's presentation which was entitled, 'Teaching Database for the Cloud'. One of the things that I really liked about Clare's presentation was that she addressed a question I was already mulling over, namely, 'how is it possible to define cloud databases?' She gave an answer that related to some NIST definitions (pdf).
A cloud database can be considered in terms of Infrastructure as a Service, Platform as a Service and Software as a Service (such as Google Spreadsheet). Depending on their task, developers will interface with different products (and different levels of abstraction). Clare pointed us to interesting sounding products such as Microsoft Azure and Oracle on Demand, neither of which I had ever heard of before.
This is obviously a rapidly changing field! If you require any further references to either cloud related papers (or products), Clare might be able to share a useful URL.
Paper Session 3 - Embedding Technology
Jackie Campbellfrom the University of Leeds began the final session by presenting a paper entitled, 'Inquiry based learning database oriented applications for computing and computing forensic students'. Two activities or tools were described. The first was an SQL Quiz application, where students were challenged to compose correct SQL statements. The second was more of a software maintenance task, where students were asked to investigate and carry out a number of fixes to an existing application developed using something called Oracle Apex.
I personally consider maintenance activities to be really valuable for a number of reasons. Firstly, maintenance is a substantial on-going challenge. Database designs are likely to be particularly affected if software applications or businesses merge. Businesses, of course, continually change and evolve (as must the software systems that they support). Understanding how (and where) to change or correct database queries may require students to navigate their way through unfamiliar systems. This, in itself, is likely to be an intellectually challenging task.
The presentation by Craig McCreath (and supported by his supervisor Petra Leimich) reminded me of an earlier presentation about the JISC WILD project that was held at the HEA mobile event a number of weeks ago. The underlying ideas were very similar: using mobile devices to elicit responses from students to (attempt) to assess their understanding of materials. Craig did a fabulous job at presenting, and I had a sense that the application he had developed for his final year project was easy to use.
The final presentation was, 'Using Video to Provide Richer Feedback to Database Assessment' by Howard Gould, Leeds Metropolitan University. Howard addresses the important questions of: 'what kind of feedback would most effectively benefit our students?, how do we make the best use of our time to provide this feedback?, and, how might we practically provide video feedback?'
Howard's paper was, in essence, a practice paper; I think this type of paper can be really valuable. One of the challenges that lecturers face is to offer effective and useful feedback on entity relationship diagrams, showing students alternative database designs. I sense that providing feedback on any kind of non-written notation is something that is intrinsically very difficult (and can include notational systems such as mathematics and music). Howard solves the problem by recording screen captures after having spent some time initially looking at a student submission. Practical challenges include file sizes (the videos themselves can be very big), and on-screen flicker (since an inexpensive digital camera is used as opposed to an expensive licence for Camtasia).
Discussion
After all the paper sessions a discussion was initiated by Alistair Monger (Southampton Solent), David Nelson (University of Sunderland) and Charles Boisuert (Sheffield Hallam).
Alistair pointed us towards a number of useful assessment resources, beginning by mentioning that the QAA code of practice requires formative assessment. He also mentioned REAP, an abbreviation for Reenginerring Assessment Practices in Higher Education, followed by TESTA, Transforming the Experience of Students through Assessment (JISC). Alistair mentioned the importance of having an audit trail of feedback, and mentioned a system called GradeMark.
David raised the perennial issue of collusion and plagiarism and made the point that assessment should always be at such a level that it is difficult for students to quickly 'find answers'. One solution might be to have time constrained assessments, perhaps in a computer lab (something that I remember from my days as a computing undergraduate), and the production of a portfolio which shows evidence of understanding.
Charles pointed us towards the idea of Nifty assessments. Charles mentioned the point that markers mark in different ways and that there will be variability. Another point was how to help those students who always seem to struggle with the subject.
The ensuing discussion took us into issues such as the importance of good feedback (and explaining the importance of why certain subjects are assessed), differences between both individual students and cohorts, and a reference to something called the Database Commons Initiative.
Summary
Although I'm not directly involved with the teaching of databases and database technologies I found this to be a very interesting event for a number of different reasons. The first is the difference in the variety of database related topics that are now taught; things are rather different from my undergraduate days when database education began (and ended) with SQL syntax and learning about different types of joins. The domain is now a lot richer than it ever was.
There are now different levels of 'cloud' databases, small embedded databases, huge data warehouses, object-oriented databases and XML databases. Other themes that might have been perhaps discussed are the connections between database teaching and software design, alerting students to issues such as making database abstraction layers and stored procedures (but perhaps these issues have been explored in earlier workshops).
The second reason also relates to this issue of richness. Software engineers and system designers are now faced with a myriad of different choices in terms of architectures (how to set up large systems) as well as products, both commercial and open source. Understanding what different products do and how they might support business objectives are issues that software professionals always need to bear in mind. These 'industrially connected' points relate to the issue of certification. The teaching and learning of databases is, perhaps, now an endeavour that is shared between academia and industry. Perhaps the role of higher education teaching and learning in this area is to provide a useful context to help students to get to grips the more practical dimension of professional certifications.
Another thought that came to mind was I felt that there was a degree of useful cross over between other HEA events, particularly the joint employability and computing forensics event that I attended. During the forensics part of this day I remember a fair amount of discussion about the sharing of 'data sets' which could be used to enable students to hone their computing forensic skills. It struck me that database technology educators are faced with a similar challenge.
A final comment is I personally consider that the subject of databases is a pretty fundamental area of computing education. I have to mention, of course, the Open University's own database course, M359 Relational databases: theory and practice.
Database education is an area is necessarily rich: students will be exposed to different languages, problem domains and wider system architectures and designs as soon as they set to work on 'real world' applications. Databases represent a domain of software technology that ultimately helps 'to get jobs done'. Where would we be without them?
TLAD (Teaching, Learning and Assessment of Databases)
I always enjoy visiting Manchester. Having spent many years there (both as an undergraduate and a postgraduate) visiting Manchester almost feels as if I'm coming home. Plus, travelling directly to the computer science building (as a computer scientist) feels as if I'm returning to a 'spiritual home'!
The reason for my most recent visit was to attend the 9th TLAD workshop, which was all about the teaching, learning and assessment of databases. Below is a summary of the event. I hope it is useful for someone.
Paper Session 1 - Data Mining
The first paper that was presented during the day was entitled 'Teaching Oracle Data Miner using Virtual Machines', by Qicheng Yu and Preeti Patel, both from London Metropolitan University, and expertly presented by Preeti.
Preeti made the point that employers are demanding up to date practical and technical skills. The issue of employability was, of course, an issue that was discussed within an earlier HEA event entitled Enhancing the Employability of Computing Students. To address the important issue of technical skills, educators are necessarily faced with the challenge of how to enable learners to make use of industrial tools and products. One of the solutions (in the area of database and data mining education) was to make use of virtual machine technology, such as the Microsoft Virtual PC, Oracle Virtualbox and a product called VMWare Player.
Some of the challenges that have to be addressed are technical i.e. how to share drives through a host computer, and financial or legal challenges, such as how to make sure that any solutions are correctly licenced. Preeti pointed us to a product known as WEKA, which I had never heard of, but it seemed that many of the audience had!
The second paper of the morning was by Hongbo Du from the University of Buckingham. His paper entitled, 'Data mining project: a critical element in teaching learning and assessment of a data mining module' won the 'best paper' prize of the workshop. Hongbo mentioned what seemed to be an important paper in the area, namely, something called the CRISP-DM standard (wikipedia), which enable students to gain an understanding of the data mining process. Hongbo also presented a general framework for assessment (within his paper) which drew upon this methodology before presenting three different case studies.
The key points that I took away from his presentation was: group work is important (but very often students don't want to do this since they might want to own their own scores completely), and that the domain of application is very important too, and represents a substantial area of complexity that students need to necessarily grapple with.
The final presentation of the first session, entitled, 'Making data warehousing accessible' was by Tony Valsamidis from the University of Greenwich. Tony began by presenting some of the problems, such as the availability of large data sets (an issue that I shall return to later), unwieldy query languages and unfamiliar domain and data models.
Tony spoke of a number of different technologies and techniques, some I had used in anger (such as Visual Basic for Applications), others I had heard of but had forgotten the meaning of (such as OLAP). An important issue that was raised was that of data sanitization: when you move data between different systems, things might not be directly compatible, so database practitioners might have to design some magic data transformations.
Paper Session 2 - Database and the Cloud
The fourth presentation was rather different. Mark Dorling from Langley Grammar School described his teaching practice and association with a project called Digital School House.
Mark trained as a primary school teacher but he is now working within the secondary sector. He described how he helps students to understand the key concepts of data, information and logical operators, sometimes applying kinaesthetic learning techniques (i.e. movement). Mark also described the use of something called independent learning videos (allowing students to remember how to use elements of applications). This reminded me of the industrial term of 'on-demand e-learning', where learners can call up screen casts or bite sized presentations about how to carry out or complete different tasks.
Mark also showed how secondary school pupils could make use of cloud application, such as Google Spreadsheet, to enable an entire class to enter data (and for the data to magically appear on an interactive whiteboard). Mark's presentation (and paper) got me thinking about how I might potentially adopt some of the pedagogic techniques that he described into my own teaching practices.
Mark's description of how he makes use of Google Spreadsheet lead us directly to Clare Stanier's presentation which was entitled, 'Teaching Database for the Cloud'. One of the things that I really liked about Clare's presentation was that she addressed a question I was already mulling over, namely, 'how is it possible to define cloud databases?' She gave an answer that related to some NIST definitions (pdf).
A cloud database can be considered in terms of Infrastructure as a Service, Platform as a Service and Software as a Service (such as Google Spreadsheet). Depending on their task, developers will interface with different products (and different levels of abstraction). Clare pointed us to interesting sounding products such as Microsoft Azure and Oracle on Demand, neither of which I had ever heard of before.
This is obviously a rapidly changing field! If you require any further references to either cloud related papers (or products), Clare might be able to share a useful URL.
Paper Session 3 - Embedding Technology
Jackie Campbellfrom the University of Leeds began the final session by presenting a paper entitled, 'Inquiry based learning database oriented applications for computing and computing forensic students'. Two activities or tools were described. The first was an SQL Quiz application, where students were challenged to compose correct SQL statements. The second was more of a software maintenance task, where students were asked to investigate and carry out a number of fixes to an existing application developed using something called Oracle Apex.
I personally consider maintenance activities to be really valuable for a number of reasons. Firstly, maintenance is a substantial on-going challenge. Database designs are likely to be particularly affected if software applications or businesses merge. Businesses, of course, continually change and evolve (as must the software systems that they support). Understanding how (and where) to change or correct database queries may require students to navigate their way through unfamiliar systems. This, in itself, is likely to be an intellectually challenging task.
The presentation by Craig McCreath (and supported by his supervisor Petra Leimich) reminded me of an earlier presentation about the JISC WILD project that was held at the HEA mobile event a number of weeks ago. The underlying ideas were very similar: using mobile devices to elicit responses from students to (attempt) to assess their understanding of materials. Craig did a fabulous job at presenting, and I had a sense that the application he had developed for his final year project was easy to use.
The final presentation was, 'Using Video to Provide Richer Feedback to Database Assessment' by Howard Gould, Leeds Metropolitan University. Howard addresses the important questions of: 'what kind of feedback would most effectively benefit our students?, how do we make the best use of our time to provide this feedback?, and, how might we practically provide video feedback?'
Howard's paper was, in essence, a practice paper; I think this type of paper can be really valuable. One of the challenges that lecturers face is to offer effective and useful feedback on entity relationship diagrams, showing students alternative database designs. I sense that providing feedback on any kind of non-written notation is something that is intrinsically very difficult (and can include notational systems such as mathematics and music). Howard solves the problem by recording screen captures after having spent some time initially looking at a student submission. Practical challenges include file sizes (the videos themselves can be very big), and on-screen flicker (since an inexpensive digital camera is used as opposed to an expensive licence for Camtasia).
Discussion
After all the paper sessions a discussion was initiated by Alistair Monger (Southampton Solent), David Nelson (University of Sunderland) and Charles Boisuert (Sheffield Hallam).
Alistair pointed us towards a number of useful assessment resources, beginning by mentioning that the QAA code of practice requires formative assessment. He also mentioned REAP, an abbreviation for Reenginerring Assessment Practices in Higher Education, followed by TESTA, Transforming the Experience of Students through Assessment (JISC). Alistair mentioned the importance of having an audit trail of feedback, and mentioned a system called GradeMark.
David raised the perennial issue of collusion and plagiarism and made the point that assessment should always be at such a level that it is difficult for students to quickly 'find answers'. One solution might be to have time constrained assessments, perhaps in a computer lab (something that I remember from my days as a computing undergraduate), and the production of a portfolio which shows evidence of understanding.
Charles pointed us towards the idea of Nifty assessments. Charles mentioned the point that markers mark in different ways and that there will be variability. Another point was how to help those students who always seem to struggle with the subject.
The ensuing discussion took us into issues such as the importance of good feedback (and explaining the importance of why certain subjects are assessed), differences between both individual students and cohorts, and a reference to something called the Database Commons Initiative.
Summary
Although I'm not directly involved with the teaching of databases and database technologies I found this to be a very interesting event for a number of different reasons. The first is the difference in the variety of database related topics that are now taught; things are rather different from my undergraduate days when database education began (and ended) with SQL syntax and learning about different types of joins. The domain is now a lot richer than it ever was.
There are now different levels of 'cloud' databases, small embedded databases, huge data warehouses, object-oriented databases and XML databases. Other themes that might have been perhaps discussed are the connections between database teaching and software design, alerting students to issues such as making database abstraction layers and stored procedures (but perhaps these issues have been explored in earlier workshops).
The second reason also relates to this issue of richness. Software engineers and system designers are now faced with a myriad of different choices in terms of architectures (how to set up large systems) as well as products, both commercial and open source. Understanding what different products do and how they might support business objectives are issues that software professionals always need to bear in mind. These 'industrially connected' points relate to the issue of certification. The teaching and learning of databases is, perhaps, now an endeavour that is shared between academia and industry. Perhaps the role of higher education teaching and learning in this area is to provide a useful context to help students to get to grips the more practical dimension of professional certifications.
Another thought that came to mind was I felt that there was a degree of useful cross over between other HEA events, particularly the joint employability and computing forensics event that I attended. During the forensics part of this day I remember a fair amount of discussion about the sharing of 'data sets' which could be used to enable students to hone their computing forensic skills. It struck me that database technology educators are faced with a similar challenge.
A final comment is I personally consider that the subject of databases is a pretty fundamental area of computing education. I have to mention, of course, the Open University's own database course, M359 Relational databases: theory and practice.
Database education is an area is necessarily rich: students will be exposed to different languages, problem domains and wider system architectures and designs as soon as they set to work on 'real world' applications. Databases represent a domain of software technology that ultimately helps 'to get jobs done'. Where would we be without them?