Dear dataTELer,

please find here the 4 Grand Challenges we created during the workshop at ARV11:

GC1: Privacy, Data Protection, Surveillance in DataTEL Privacy

The DataTEL research must address issues with respect to data protection (or other relevant) legislation compliance, concerns with respect to individual privacy, as well as problems arising from surveillance* (social sorting, cumulative disadvantages).

1. We need to develop a new vocabulary to discuss the problems associated with privacy, data protection and surveillance. For example, how do you describe the affect that statistical disclosures have on individuals who have not revealed personal information. What are better terms to express concepts like ownership and control, when in digital systems replication and distribution is so easy that the concepts have no traction. (1 - 3 years)
2. Research is needed on how existing privacy and transparency solutions can be integrated in dataTEL practice. Further, research is desirable on how state of the art security solutions can be used to secure large datasets. (1-3 years)
3. There is a need for data awareness education for society. Such an educational program should not be limited to teaching individuals when to reveal or conceal their data, but also to increase their awareness with respect to large datasets, surveillance practices, and related problems. (1-3 years)
4. The issues around privacy, data protection, and surveillance need to be addressed from the beginning of the research and not as an add-on. Methodologies and guidelines that support this vision need to be developed to support privacy and ethical practices. (3-5 years)
5. There needs to be research on how to bridge between dataTEL researchers and ethical boards with respect to advances in technologies and research and the related privacy, data protection, and surveillance concerns that arise with them. (3-5 years)
6. User and stakeholder studies (case studies) are necessary to understand the complex requirements with respect to privacy, data protection, surveillance in dataTEL. (1-3 years)
7. Policies have to be defined to avoid unethical data mining research. (3-5 years)

Funding: LLP / FP7 funding

*the use of the word surveillance refers to the process which individualizes each member of the population (or a group), and permits the observation and recording of each individual’s activities, then collates these individual observations across the population. From these conglomerated observations, statistical norms are produced relating to any of a multitude of characteristics. These norms are then applied back to the subjected individuals, who are categorized and perhaps acted upon, either with gratification or punishment, according to their relation to the produced norm. (Phillips, Privacy Policy and PETs, 2004)

GC 2: Reduce drop out rate in online learning by 10% employing recommender systems for learning

Tasks:

Develop a set of indicators or a chain of indicators to link recommendation to drop-out rates Customize existing recommendation algos for learning,  employ recommender systems in real-life scenarios.

Time and measurement:
Time frame: 2-3 years, building on existing dataTEL work Outcome: improve the learning outcome

GC 3: TEL Dataset Grand Challenge
Define and promote a common generic infrastructure for sharing, analyzing and reusing learning resources and learning activity logs. Benefits improving learning experiences with personalization optimization of learning processes speeding up creation of new resources. Incentives Funding. We think it is more likely that a standard will be enforced by governmental bodies such as NSF and the EC (and yes, most likely there will be American and European challenges)

Stakeholders:
LMS producers, content providers, teachers

Timeline:
Anything between tomorrow and within 10 years. For learning resources there are already standards like LOM and Dublin Core. For learning activities it's more complicated (apart from very generic formats such as XML – which does not guarantee that data can be reused). Necessary steps – Issues to be solved data ownership privacy a body of accepted analysis methods, methods of research Success indicators, Quality and quantity of data in Format X. We did not agree on a scenario before Format X will appear (will there be competition, consensus, or analytical paths to such a format)

Funding bodies:
Governments, companies (Microsoft, IBM)

GC 4: ACTUALLY help students and teachers in TEL using recommender systems

Activities:

Make real time/running environments available as test applications (i.e. dynamic data sets) Identify algorithms and map them to data sets and purposes Find measures to evaluate which might include: the increase of effectiveness of learning processes the increase of efficiency the increase of satisfaction

Time: 3-5 years

datatel, grand challenge, data, arv11, datatel11, stellarnet, research agenda

Last updated 408 days ago by Hendrik Drachsler