Abstract: This thesis examines whether it is possible to predict waiting times at traffic lights in order to use that time for playing media, as for example videosequences, the reading of texts etc., which is automatically adapted to the length of the waiting period.
An adaptive algorithm is introduced which detects traffic light areas by using GPS data and calculates the excepted length of the next waiting period. The developed algorithm is achieved through the use of server client architecture. The collected information about the positions of traffic light areas and the associated delays are stored on a central database server that is accessed via an application server. The client was conceived for the use on mobile phones to guarantee the highest possible degree in mobility. The developed system is evaluated through the use of test runs, which checks whether traffic areas are correctly identified and allocated, and whether the estimated waiting time conforms to the actual wait-ing time at a traffic light. The results have been mainly positive. The traffic light zones were correctly formed and recognised. And although the amount in information for the waiting time is still very low in the traffic light zones, most calculated waiting times with the actual ones conforms. The analysis of an on-line survey regarding to preferential entertainment forms and potential subjects for the use in such microbreaks is discussed. An evaluation of the acceptance of this concept has shown that 70% of respondents would not be reluctant to using such a system.
(Fabian Schulte, Bachelor Thesis)
Abstract: Many public display systems provide interface interaction technologies that mostly face unique controller devices. Furthermore personal gadgets like mobile phones be-come more and more ubiquitous in everyday life. The question is how a system can be designed that is easy to use, works instantly on the one hand and on the other hand can interact in real-time and reliable with the display. I will explain the differ-ent connection types with its pros and cons and the impact with public user study.
(Lars Schaaf, Bachelor Thesis)
Abstract: The target of the thesis is the development of an authentication system, which uses gaze-based password entry and graphical passwords. The system should reduce the possibility for shoulder surfing attack, when using public terminals. It should also increase the password memorability by replacing the exact password recall(PIN) by recognition of picture parts. The system provides two different authentication approaches. The first one uses a normal picture on which the password is represented by circle areas. In the second one the picture is divided by a grid and the user can choose grid parts, which are representing his password. After successful system development a user study was conducted to prove the system usability. In the pre-study a small group of people was asked to select a password of their own, using both authentication approaches. The objective of this pre-study was to select two suitable pictures for the graphical passwords. Afterwards 13 persons were invited to take part in a two-day study. On the first day the participants were asked to enroll in the system and immediately afterwards to login one time. Two days later the participants were asked to login again, in order to check how good they were able to recall their passwords.
(Yordan Terziev, Bachelor Thesis)
Abstract: Advertising companies are always looking for new and innovative prospects to present
their products and offers to costumers. Often local advertisements disappear in
the mass of advertising or do not reach the desired target group.
The goal of this work was to develop a new advertising system which enables advertising
companies a new opportunity to present their ads. The user can interact with
the system and can choose the offers which are interesting for him. The system is location
based and presents the nearest local companies’ ads to foreground as well as
their distance to the taxi. The work is done in cooperation with TaxiMedia. Therefore
we developed a System for a cab. First of all we compare similar systems and discuss
the different views of the stakeholders. The core design decisions and main advantages
will be described afterwards. Thereupon the architecture of the howl system
and the development of the cab‐client in detail will be explained.
(Max Pfeiffer, Studienprojekt)
Abstract: In this project, an approach for implicit interaction with web advertisements using eye tracking is introduced. The approach uses the gaze data collected from the user to determine the user’s interest. With the collected data, the system shows the users the advertisements that match their interest.
This thesis describes the conceptual approach and analyzes the related work for the topic of eye tracking based interaction with web content. Based on the documented approach, a system architecture and implementation is described.
The thesis explains the user study which was used to evaluate the implemented sys‐tem and analyzes its results.
Finally, a conclusion to the approach is given and further work is investigated.
(Dirk Kuschnerus, Studienprojekt)
Abstract: In dieser Arbeit wird das dynamische Abspielen eines Werbevideos mittels eines Eye-
Trackers mit dem statischen Abspielen eines Werbevideos verglichen. Dabei wird untersucht,
ob der Benutzer die dynamische Abspielfunktion durch die Verwendung eines
Eye-Trackers erkennt, und welche Art des Abspielens eine höhere Aufmerksamkeit erzeugt.
Dazu wird die Webseite www.Amazon.de um ein zusätzliches Frame mit einem
Werbevideo erweitert, welches sowohl die statische als auch die dynamische Abspielfunktion
ermöglicht. Durch eine Benutzerstudie wird ermittelt, welche Abspielart mehr
Aufmerksamkeit erzeugt. Dazu wird sowohl die Dauer gemessen, wie lange der Benutzer
die Werbevideos in der jeweiligen Abspielart betrachtet, als auch an wie viele in den
Werbevideos gezeigten Produkte und Marken sich der Benutzer erinnern kann.
(Achim Brandes, Studienprojekt)
Abstract: Until now existing advertisement methods mostly use static information to deliver advertisements to a customer. Context‐aware advertisement promises to achieve better tailored messages and contents adapted to the user’s situation, environment and personal interests. But context‐aware advertisement was only successful in internet services and pages. Efforts to push this new advertisement method into the mobile environments of users seem not to have significant success. This work introduces a new concept of a context‐aware advertisement platform, which allows a dynamic distribution of advertisement messages without being focused on a particular target system. The result is a multi‐component prototype, which can be deployed as a middleware between the advertising companies and customers.
(Christoph Evers, Master Thesis)