#digitalmemorytraces

Platform Design and Development of Digital Memory Traces - Applying our use case driven approach, we designed and built the Digital Memory Traces platform. It offers functionalities to support the definition and development of service ideas by addressing media collection and use as well as social interactions. The definition of different media types and functionalities to participate and engage is key to targeting a wide user community. The media types used during intake and developed throughout the project emanate from our model. Applying this model, we identified the following formats: Story, Memory Fragment, Reflection, Statement, and Photo. To capture the experiences of sharing memories, we introduced an additional format: Experience. Our platform applies storytelling and encourages a close look and social interaction, which is solely created in a co-creation manner.

The discussion in the preceding pages has confirmed the need for web-based applications that facilitate participation, creation, sharing, and co-creation of personal content-based information services. We refer to these web-based applications as media-sharing platforms. The literature focused on the micro level – the individual creating, capturing, storing, and sharing personal content intelligence. Business services need stories, with people in them, to engage, interact, and motivate clients. Therefore, at a macro level, we need to rethink how to facilitate mass creation, sharing, and embedding of stories, photo collections, reflections, goals, and 'memories'.

User Interface Design

The focus is on students and teachers-developers, thus promoting change in teaching and learning, contributing to improving the learning process. The user can access DMT, using the functionalities that allow exploring the traces, visualizing results obtained from parameters defined in a trace, starting the configuration of a trace, and obtaining snippets using the parameter estimation process. Snippets are pieces of multimedia that the machine learning algorithm selects for human annotation in order to help it learn how to spot a specific category. They include segments of videos and are painful interactive processes of data collection where human annotation is required. At the moment, the application does not have the used data stored in an adequate way, constituting future work. User access to the application, except for adding new algorithms and/or editing estimates, is conditioned on authorization.

Based on the study of similar applications and user experience design principles, it explains how the visual design should be, underlining the importance of brevity, or the user's ability to interact with less content, making extensive use of the horizontal scroll and load more. This is very useful for educational settings, where it is difficult for a person to interact with a great amount of simultaneous information. It suggests research on the design of data quality measures with an ergonomic approach and using them in the user interface, between machine learning researchers and the audience they seek to serve, in the context of machine learning for video data. Machine learning research has been primarily concerned with accuracy, or minimizing the difference between algorithm and human labeling assignments. They suggest a positive approach, assuming that they can design sampling schemes, factors that affect performance, and conditions that will improve sample annotation compliance.

Technical Architecture

The platform can be seen as two different layers. The lowest one is the storage, and the upper one joins storage and preservation. Under the preservation layer, creation and some specific retrieval actions are added. This approach finishes the basic concept framework of the proposed platform. Fine-tuning of the technical architecture incorporates a number of specific requirements for the platform. The first basic requirement is the digital preservation of creative, intellectual human activities that include the representation of these activities in a large number of different ways. There are basically two approaches to cope with this problem: the use of any additional metadata, which helps to describe every memory and its meaning in a highly accurate way, and the development of memory infrastructures that benefit from information redundancy and a high degree of tolerance even when part of this metadata is missing, noisy, or just wrong. These may be added to the platform as they become stable and compatible, provided they satisfy platform flexibility requirements.

In addition, another requirement must be added to allow for easy and highly intuitive access to the platform's contents. However, the search engines at this level are limited to exploring contents, not meaning and interpretation. For the same heavy computer costs, analyzing abstract contents is expected to be supported by professional expert systems available in a third-party environment as plugins. This will bring additional difficulties to the integration task, since some memory contents may be linked to specific professional analyses providing specific meaning. Moreover, the application of these professional expert systems is just complementary with respect to the memory contents distribution and availability. For these basic needs, an easy-to-access service layer is provided. It is aimed at replacing standard repository accesses with plug-in explorative services. These professional services access the data storage level using both stored memory contents and an extended semantic metadata tool and data infrastructure.

User Experience Considerations

Consideration of the user experience during the design of DMT was an ongoing and underlying theme. Trustworthiness develops as a more understated and latent aspect of educational technology. We understand our technology is a means that students can use for learning and that this raises the issue of how to design technology so that our users can trust it. Trustworthiness is related to our comprehension of the integrity of the information, and it is important because students rely on the technology to deliver accurate representations. A trustworthy system does not make itself known; it is something that is achieved through human-computer interaction. Trustworthiness is related to prolonged, active, and intense mental activities, motivational inertia, and satisfaction. The contributions to trustworthiness of these shortcuts are as follows: if a user has a good experience when they use an information system, then the user will be confident in the capabilities of the system. This is true whatever the design philosophy, and if certain user expectations are not met, the level to which a user trusts the system decreases.

Researchers must draw on standard methods and guidelines for keeping the users' interest; otherwise, it can lead to the loss of the users' trust in using the system or having faith in the choices the system has made for them. Factors to be taken into account involve component arrangement and format, timing of programs, upload time, protocol consistency, high quality, maintaining the aesthetics of design, minimizing cognitive overload, and minimizing human effort. Overall, taking these factors into consideration enhances the attractiveness and comprehension of trustworthiness for the user, which can then result in the user being confident when they use a decision support system, and it can also provide satisfaction to their needs. There will be trust in the system and consequently satisfaction. However, an undersupported, low-readiness system is just the opposite. The user's levels of trust, confidence, and satisfaction will drop, and both goals and the potential that technology can afford decrease. The alliances that have been built with the repositories and their responses to the study are perceived as ongoing and not just for the students.