PixMeAwayTravel recommendation platform
ecosioSolutions for the exchange of electronic documents
E-Business and E-Commerce include, besides other influencing domains, rapid changes to economic and social processes. For this reason, the E-Commerce group pursues taking an encompassing view at such processes reaching from analyzing enterprises and their respective processes to evaluating the entire market, as well as is active in creating ideas and solutions for meeting the demands of the respective stakeholders.
Overall, the E-Commerce group performs research and creates solutions in the Business-2-Business (B2B) and Business-2-Consumer (B2C) domains. In the B2B domain, the group focuses on (i) business (process) modeling and definition and consequently the specification and implementation of e-business systems while considering and contributing to semantic web research and service-oriented architectures, (ii) ontology engineering and information integration, (iii) "Web Science" focusing on network analysis and content as well as text mining. In the B2C domain, the E-Commerce group performs research targeting (i) visual interaction paradigms as well as (ii) mobile applications. Thereby, our group focuses especially on the development and application of our research in the area of e-tourism and its mobile applications.
Parts of the EC group are working on systems based research, namely on virtualized HPC systems, massive-scale data analytics and energy efficient ultra-scale distributed systems. Our goal thereby is to develop speculative methodologies and techniques to allocate resources in an energy efficient way, but at the same time meet requirements of the end users, like the certain latency of the applications. Our approaches are evaluated quantitatively on different scales by simulations, benchmarks or on experimental clusters. We strive to develop generic methods and methodologies but we let our research be inspired by a set of real world applications from different domains like life science (e.g., DNA sequencing) or social media (e.g., Facebook size data centers).
The Slovak University of Technology in Bratislava has awarded the medal of Johann Wolfgang von Kempelen to Prof. Tjoa for his significant contribution to advancement of knowledge and collaboration in Informatics and Information Technologies.
The Startup Pixtri from the Vienna University of Technology – E-Commerce Group won the 1st place at the ENTER Conference 2013 for its significant contribution to the growth and success of technology adoption, integration, or use in the tourism industry.
Based on extensive scientific and practical preparatory work, the core functionality of PixMeAway comprises an image-based search and recommendation engine in the tourism sector.
Imagine you want to go on vacation, but yet, you do not know exactly where you want to go or what kind of trip you want to take? In such a case you might query different search engines in the Web by entering text-based search criteria. But what should you enter, if you do not know your travel desires in advance? It follows, that a part of such “un-decided” users enter rough keywords of their trip imaginations, which results in a big amount of “non-classified” information. The other part of these users consults offline tourist agencies to get inspired by professionals. Those professionals mostly work with pictures of travel destinations to find out the hidden desires of their customers in order to be able to suggest a trip. PixMeAway does almost the same online, finds out the user’s travel type via pictures and suggests point of interests, activities, tours or regions.
The platform www.pixmeaway.com works as follows. A user selects the images that particularly attract him or inspire him. The system calculates a certain “tourist matrix” of the user, which comprises seven pre-defined travel factors (these factors are a mixture of travel types and personality types based on scientific standards). By the use of this profile the system recommends appropriate touristic objects.
Technically, the solution is based on a sophisticated recommender system in combination with an administrative backend interface for Points of Interest (places with high tourist attraction) and a module for travel planning. Furthermore, the platform incorporates aspects from social media communities to share trips/destinations/activities/etc. with other users.
Our EC Group member, László Grad-Gyenge, was awarded for his innovation. He won the first place in Innovation category on eFestival 2012, Hungary.
He created a lot nicer and more secure CAPTCHAs than the currently existing ones on the market by launching an animated version. He offers his CAPTCHAs - under certain limits - as a free web service on the website of the product: http://www.hellocaptcha.com/. He was also interviewed on the main radio station of Hungary about his innovation. His first client in Austria is the Austrian Parliament, which uses the HelloCaptcha for the protection of the online consultations. He is now moving to research social networks within the EC Group to progress with his innovations in this rising new area.
We are pleased to introduce the new member of the eCommecre group - László Grad-Gyenge. László studied Computer Science at the Eötvös Loránd University, Budapest, Hungary. He has experience in the following areas: global optimization (iterative methods), computer vision, natural language processing. His next mission is to study the scale free networks in the social space, and to analyze memes.
László studied computer science at the Eötvös Loránd University, Hungary. He wrote his MSc thesis in the area of global optimization, using iterative methods. He was trying to find good starting points for local searches using function approximators to estimate the space, based on the sample data collected by local searches. He also tried automatic feature extraction to help the function approximators. He finished it in 2001.
To continue his MSc studies, László started a PhD at the same university, in the same area. He applied his search methods in the following areas: satisfiability, geon space, facial expression recognition, exploring the visual cortex of the primates.
Next to his PhD he started to build his own company (www.creo.hu). The primary fields were business application development and web development. For example: http://www.finalexam.eu - an automated text examination system, http://nrk.nive.hu - Europass issuing and adminstration system. He also had innovative projects, that are some of the products of his company, i.e.: www.simplefont.com, www.hellocaptcha.com. His company took his time, so he resigned his PhD status and was building the company for approximately 10 years.
In 2011 he decided to restart his PhD, so László went back to his university, and started to work in the area of computer vision with saliency and eye direction estimation. It was a short story, because after two months his supervisor asked him to switch to the area of natural language processing, and to create a question answering tool, by extracting FAQ information from dialogs. He worked two semesters in the latter area. In the first semester he created a framework called nBoss for web applications in J2EE. The second semester he spent at the University of Vienna as an Erasmus student and built the theoretical background for textual knowledge representation.
As a PhD student László learnt to know Vienna, and liked it very much, so he decided to stay here, if possible. So he applied for the the doctoral college "Adaptive Distributed Systems", he got approved and started to work here in 2012. His next area will be network research. László is planning to study scale free networks and to analyze memes.