i am having a short break from TU wien. i will return in fall 2024, probably with all-new topics then. for now, this place is closed.

Compared to online systems for collaborative authoring like the google office suite or miro, collaborative annotations of and communication about online documents is still quite underdeveloped. Medium tries to do a some highlight-sharing and inline annotation stuff which is certainly interesting, but maybe there is more.

If we consider reading documents not as a passive consumption of text, but an active process of structuring, organization and annotation (like it would be i a learning context), even more possibilities open up.

This master thesis for one or two candidates tries to develop a prototype of a system where readers collaboratively work with a given document with the goal to not only understand it, but enhance it with eg. current events, connotations, crosslinks to other materials, links into the internet etc.

Ultimately, the goal is to create an environment where the barriers between the roles of teachers and students get torn down, and both groups can benefit from each other.

since the beginning of the scientific revolution, the publication of information – insights, knowledge, theories, etc. – was one of the main driving forces of science. it can be argued that the scientific revolution was fuelled by a rebellion of newton’s contemporaries against secrecy and private profiteering as much as a different way of thinking.

for centuries, the format of the printed scientific article has been refined into a number of different standardised layouts (dependent eg. on the branch of science, or the publishing medium) and conventions (eg. for citations, or for the order of authors. then, the internet came along and, over night, made that all obsolete — only that it didn’t. scientific publishing is still defined by the format of the printed paper, creating a number of problems in the digital.

it is time to rethink scientific publishing formats.

in a digital-first publication format, references could not only refer to other work, they could refer to the very sentences and words of a quoted article; the multitude of reading devices, from smart phones to 4k screens and beyond, deserve optimised presentations of text and images depending on the strengths of each screen; printed versions could flow into the various standard paper sizes; and since scientific work often gets refined, revised, and discussed, we could even think about ways to reflect this in a new format.

the goal of this diploma thesis (and beyond?) is to define a new standard digital-first scientific publication format, based on the affordances and strength of digital media on one hand, and on the needs and ideas of scientific writers. previous attempts to do this should be documented. also, a renderer for LaTeX (or similar) should be implemented that translates written work into this new format.

the goal of this bachelor or master thesis for one or more students is to explore the field of distributed microphones.

as the pandemic has shown, picking up audio from a room full of people (eg. for hybrid meetings) is quite a challenge. room microphones only work properly in rooms that have good audio in the first place (eg. no reverb) which often cannot be guaranteed. a promising approach to solve this problem is to use the many network-connected devices already present in a room full of people in order to create a singular audio-stream (for transmission and/or recording) that could potentially be of much higher quality than any other solution.

one or more prototypes of such a solution should be implemented and tested, based on different approaches of control (automatic control vs. user-control, push-to-talk vs. »throw the ball«, etc).

the CMU mocap dataset is a collection of 2537 actions like walking, talking, jumping etc. generated from motion capture. it is used, among others, in machine learning settings, and can there become a source of (well-hidden) bias eg. if all those motions were only generated from male, or healthy, or typically western people. the goal of this master thesis is to discuss how bias is defined in the context of such a dataset, how the dataset can be evaluated for various biases, and conduct exemplary evaluations for one or more bias dimensions. data set bias evaluation is an exciting new research area that is steadily gaining in importance.

Speaking in front of a real audience is a huge challenge for some people. The goal of this master thesis for one person is to create a VR system that lets you experience talking in front of a large crowd. It should include a 360° recording of a lecture hall or similar setting, the possibility to include your own slides as the presentation shown on the projection, and random events like audience showing up late or leaving, dropping pens, talking, etc. Ideally, it ends with a round of applause at the end. The system is to be implemented and evaluated.

werkzeuge, insbesondere digitale werkzeuge, sind bestimmend dafür, was wir tun können. wir beobachten einen klaren trend weg von traditionellen »computern«, hin zu allerlei devices, die oft mit anderen interaktionsmodi als maus und tastatur bedient werden: multitouch, sprache, gesten, etc.

es wird aber immer offensichtlicher, dass vor allem multitouch gegenüber der »traditionellen« maus+tastatur-interaktion von desktop.-betriebssystemen trotz aller refinements einige wesentliche definizte hat. in dieser diplomarbeit für 1 person soll untersucht werden, welche gestaltungspotentiale eine gemischte multitouch/mouse interaktion mit sich bringen würde. es wäre ein exploratives system (zb. grafik-editor) zu erstellen, in dem die gleichzeitige und synergetische nutzung mehrerer interaktionsformen (maus, pen/cil, multitouch, tastatur) experimentell untersucht werden kann.

In Denkweisen der Informatik (ways of thinking in informatics), a unique first-semester course for all informatics bachelor students at TU Wien, we organized (more or less) weekly panel discussions where experts of the respective fields answer and discuss the students’ questions. To collect the questions, we use the Q&A-Tool of the zoom webinar format.

While this is readily available, it has some serious drawbacks, eg. questions can only be submitted once the webinar has started, while it would be helpful if students could post questions whenever they have them.

The goal of this master thesis project is to design and implement a little platform (probably using django) that mirrors the possibilities of the Zoom Q&A-tool and extends them in the direction we need it. There are some role models, like mentimeter, mukana or the old best and worst of informatics tool we used when Aurora was still around, but basically this should be designed from ground up along the necessities of its intended use.

A format to translate lecture slides into a semi-booklike format has been developed and refined over the recent years. An example can be seen here:

https://dwi.lva.tuwien.ac.at/dwi20221117.html

Slidebooks are created using a convoluted pipeline of processing steps, from the creation of slides in Apple Keynote, the annotation of these slides using the »presenter notes« functionality in Keynote, the extraction of all content from Keynote using AppleScript, ImageMagick and curl and the encoding into a text-only format, to a python script that creates an HTML + javascript-file that can then be uploaded to a web server.

This pipeline is not only relatively slow, unreliable and error-prone, it is also a nightmare to maintain and update. The goal of this master thesis for one person is to take everything that has been learned so far, completely rethink the process and create a »service« that makes it easy to create slideBooks.

In addition, new elements for SlideBooks that enhance the interactivity (like quizzes, crosslinks between slidebooks, gameful elements) should be devised, designed and prototyped or implemented.

ChatGPT has created a conundrum for teachers everywhere. Simple »text production« tasks are endangered by the glorious »stochastic parroting« qualities of chatGPT, and asking students/pupils to summarize something will soon be a completely useless assignment.

So discussions, how this can be avoided/mitigated/used productively have sprung up everywhere. Some US school districts are banning chatGPT, detection tools are being developed, and others try to find productive uses of the system in classes - and this is just a short overview over some reactions. As chatGPT can also produce quite competent code, these problems affect not only text-centered assignments, but also introductory coding or engineering classes.

In this bachelor thesis (in german or english language) for one or two students, the goal is to gain an overview over the current discussion, and develop mitigation strategies for first-semester university courses with 500+ students (like we have at TU Wien), especially in informatics-programs

smart phones are amazing collections of sensor technologies. can these sensors, especially location, orientation and maybe the camera be used to facilitate new forms of teaching and learning in the course »ways of thinking in informatics« (denkweisen der informatik).

the goal of this bachelor or master thesis is to find a simple and easy to adapt strategie that can be used to create »distribued« learning units. this includes looking at existing approaches like educaching, QR rallys or hybrid scavenger hunts,, designing and implementing elements of an approach that is easy to maintain and fill with content, and evaluate those elements with students in the context of a large first-year informatics course.

cory doctorow is an science fiction/tech fiction author creating books in amazing frequency. some of his books are set in a near future where the technologies seem absolutely realistic or at least almost possible. this diploma thesis will explores how realistic those the technologies really are? using some of his works as a starting point (esp. the trilogy little brother, homeland and attack surface as well as walkaway and the novella-collection radicalized) the work should examine how much of the technology described in a »near future« settings are attainable or available today. if they don’t exist, could they be implemented/created? should they exist? the work should pose and answer technological, ethical and political questions regarding the concepts of his books. ideally, a particularily interesting technology is implemented in a prototypical way.

zitate werden in videos gerne mit einer anderen stimme vorgetragen, also nicht von der bzw. vom vortragenden, sondern von einer passenden anderen stimme. zwei beispiele dafür finden sich in einem video von jan böhmermanns ZDF magazin royale vom 27.5.2002:

• https://youtu.be/Xdm8SG8_v0I?t=1202

• https://youtu.be/Xdm8SG8_v0I?t=1228

leider gibt es keine software (konkret für macOS), mit der man das visuell anspruchsvoll, aber auch einigermaßen einfach automatisiert als video produzieren kann. dabei sind die requirements hier wirklich klar definiert.

im rahmen einer bachelorarbeit für eine oder zwei personen soll eine software auf macOS implementiert werden, mit der das einfach möglich gemacht werden soll. neben der einfachen bedienung und dem export als video steht dabei die qualität des gesprochenen texts im vordergrund. dafür sollen existierende services mit hervorragender qualität (zb. von amazon oder google) genutzt werden. entsprechend ist eine evaluierung dieser services (qualität, kosten, verfügbarkeit) teil der bachelorarbeit.

ein wesentliches element der software ist weiters, dass man das erscheinungsbild des textes detailliert definieren kann, um den visuellen anforderungen der videoproduktion gerecht werden zu können. über eine mögliche kommerzielle verwertung kann gerne auch nachgedacht werden.