Research activities

The Department of Information Technology is or has been involved in the following main research areas:



Current research interests and projects


Virtual Reality and Multimedia

One of the main research fields of DIT is virtual reality and multimedia. In this field DIT has been recently involved in several European IST Projects: PISTE, ARCO, INTUITION, MM4All, Web2.0ERC as well as direct collaboration with other universities and major industry partners. PUE is one of founding members of the EuroVR Association.

Main research achievements of DIT in the area of virtual reality and multimedia is development of the Flex-VR approach and the ARISM approach.

Difficult creation of meaningful complex interactive 3D content is one of the main factors which limit wider use of interactive 3D applications. The Flex-VR approach enables creation of configurable virtual reality applications, i.e. applications in which content can be relatively easily configured (created or modified) even by inexperienced users. The Flex-VR approach can be applied to building local as well as network-based VR applications. The Flex-VR approach enables building virtual reality applications in fields of e-education, e commerce (e.g., virtual shops), electronic guides, customer support, scientific visualization, geographical information systems, interactive television, etc. The Flex-VR approach has been used and further developed in three international projects: PISTE, ARCO, Periscope and a national project iTVP. The technology has been partially patented at the European level.

ARISM – Augmented Reality Interactive Scenario Modeling – is a new approach to creation of interactive educational scenarios in mixed reality environments. In this approach, the process of building augmented reality learning environments is divided into three stages: content design, content creation, and content setup, each of them performed by users with different technical and domain knowledge. The concepts of AR-Class and AR-Object are proposed extending the object-oriented paradigm to observe the requirements of AR. Visual and behavioral features of AR-Objects are established during their creation by teachers by setting appropriate property values.

DIT research deals with the problem of data security in modern 3D virtual environments which are behavioral, interactive, user-created, massively used, and highly dynamic. Traditional coarse-grained and geometry-centric privilege modeling methods are not sufficient for such environments. Therefore, new methods have been developed at DIT which do not impose too many restrictions in the phase of creation of users’ virtual objects, because the goal of those environments is to promote user creativity and sociability. Protection concerns not only geometrical models, their relationships and structure, but also inter-object behavioral interactions. In particular, a new method called Selective Semantic Modeling (SSM) has been developed which consists of two elements: Virtual Reality Privilege Representation (VR-PR) and a Knowledgebase of Objects Behavior (KBOB). The SSM method enables privileges modeling for virtual environment behavioral objects with respect to their semantics. The SSM method is based on the concept of semantic operations, which are generated at run-time from the virtual environment’s current data model and are applicable to the access control model as a part of a privilege. Semantic consistency of the privilege set is forced by a two-phase regeneration and validation mechanism so that user privileges can still be expressed in a precise, semantically accurate and flexible way.

Another DIT research effort is an answer to unprecedented explosion of the content available on the Internet and the need for efficient methods for finding the information with high both precision and recall. At DIT, research is conducted on a new method of searching for information resources related by themes, space and time called TSTSM. The major component of this method is resource similarity measure called TST. The TST measure is compound of three measures: semantic similarity, distance in time and distance in space. The TST measure is used to assess the relevance of an information resource to a user query. In the TST method semantics, time and space are considered simultaneously. Such approach makes the method suitable for analyzing resources that are changing in time or moving. By means of fuzzy sets theory, the TST method can handle imprecise data very often found in time and geographic properties of objects. The search results are presented in 3D on a map and in the form of a ranking list. An important feature of the TST method is its independence from the structure of the underlying knowledgebase, which allows easy incorporation of the existing data sets without the need for adapting them to the TST method.

Projects in the area of virtual reality and multimedia:



Virtual Organizations

IT-SOA

Research on virtual organizations is conducted in DIT due to their importance for electronic knowledge based economy. The main research goal is to develop new models, tools and methods to support collaboration of organizations in a networked manner.

A Virtual Organization (VO) is defined as an operational structure consisting of different organizational entities and created for a specific business purpose, to address a specific business opportunity. Based on the above concept of VO, the concept of Virtual Organization Breeding Environment (VOBE, sometimes shortened to VBE) has been proposed in the literature as “an association of organizations and their related supporting institutions, adhering to a base long term cooperation agreement, and adoption of common operating principles and infrastructures, with the main goal of increasing their preparedness towards collaboration in potential Virtual Organizations (VO)”. The main aims of VOBEs are: establishment of mutual trust among organizations to facilitate their collaboration in VOs, reduction of cost and time to find suitable partners for a particular VO, assistance in VO creation including reaching agreement between partners, and VO re-configuration aiming at adaptation to new business challenges and opportunities.

While VOBEs are implemented in an ad hoc manner, research in DIT is focused on SOVOBEs, i.e., Service-Oriented Virtual Organization Breeding Environments, which are organized in a systematic way around the concept of a service. In SOVOBEs, services may be Web services, potentially integrated by an Enterprise Service Bus (ESB), as well as services performed by humans (organizations).

The research on SOVOBEs conducted in DIT aims at three specific goals. First, developing a novel architecture and a reference implementation of software platform for SOVOBEs based on the Service-Oriented Architecture (SOA), taking advantage of concepts underlying SOA for a better support for virtual organizations, such as:

  • service reuse – a given organization may provide the same service within many VOs;
  • service abstraction –the details of the implementation of services offered by a given organization within a VO are usually hidden for other organizations, because the implementation of the core business services is associated with the know-how capital that gives the organization a business advantage over competitive organizations;
  • service discoverability – services provided by organizations in a SOVOBE are described, so that both services and associated organizations may be identified as suitable for a given business opportunity;
  • service composition – a complex service provided by a VO is a result of composition of services provided by VO partners and by the SOVOBE.

Second, developing models and algorithms for the creation of VOs in a SOVOBE. A SOVOBE provides its members with a service supporting searching for partners and services suitable for a particular cooperation process. The selection of partners and services on the business level is a complex task that can hardly be automated. It encompasses the following aspects: the determination of requirements by VO stakeholders; the determination of particular requirements for VO partners and their business services and processes offered to the VO clients; the identification of SOVOBE members able to play a particular role in a business process or fulfill a particular task, the negotiation and settlement of cooperation rules and conditions, the analysis of possible VO variants in terms of conformance to requirements and efficiency of cooperation.

Third, developing models and algorithms for advanced support for agile collaboration among organizations within a SOVOBE. Currently, the collaboration among organizations is structured with process models which capture the structure of interactions among persons, organizations and software entities. The social context within which a given process model may be instantiated is not a part of the process model. In DIT a concept of social protocol has been proposed as an attempt to link process models to the social context within which the processes are instantiated. The social context is significant for VO agility. A VO is said to be agile if it can rapidly and cost-effectively adapt the way it is functioning to changes. The adaptation of a VO to new conditions can be simplified by using the social context of VO partners. Adaptation to changes often requires new resources. In a situation when performance of a VO partner suddenly decreases, a possible solution is to search for another organization, which may substitute or complement the current VO partner. The social context of VO partners may be a good area to search first.

Projects in the area of virtual organizations:



e-Governance

The research goal in this area is to develop a new approach to modeling and execution of administrative procedures and transactions. Traditional approach to process modeling assumes creating models based on operational practices; i.e., a process analyst asks employees what tasks they carry out and how. This information is then used to create process models. The same approach has been used so far to create models of procedures and transactions carried out by public agencies. Meanwhile, the public agencies have a complete different nature than commercial businesses. All activities conducted by public agencies are determined by law; thus the form of administrative procedures and transactions results directly from legal rules and regulations which are the primary sources defining the way public agencies operate.

In a typical situation, the course of a specific administrative procedure or transaction is influenced by several legal acts which can be divided into two main groups. The first group consists of general acts; for example, public administrative code which defines general rules, common for all procedures and transactions regardless of the nature and essence of a case the given procedure or transaction applies to. These general acts are superior to acts of the second group. The acts of the second group have much more detailed character and define proceeding rules for specific classes of cases. For example, building code regulates details of granting building permissions. If the administrative procedure or transaction is modeled based on these two groups of legal acts, such model will not include additional options which are correct from the legal point of view but which are regulated by legal acts not directly related either to administrative transactions generally or the specific class of cases particularly. The example here is the civil code regulating the issue of granting a power of attorney.

The research conducted in DIT in the field of e-governance aims at two specific goals:

  • Developing an approach for building administrative procedures and transactions models which is adjusted to the hierarchical nature of legislative system;
  • Developing a method for dynamic adaptation of an execution course of administrative procedures and transactions based on legal aspect appearing during the execution.

Projects in the area of e-governance:



Internet of Things and its Applications


Semantic Internet-of-Things

Recently, a trend is observed of mass usage of small yet “intelligent” devices, ready to serve in small everyday tasks. The devices are connected to the network to enable communication among them and with the environment; however, they are invisible for the humans unless certain functionality is required. Then, a selected device is automatically activated, without direct human intervention, to fulfill predefined task.

While developing IoT systems and applications, two basic problem are observed related to human-device interfaces (a cooperation), and to programming devices’ behavior dynamically, in action-driven, possibly ad-hoc mode. So far, attempts to solve these problems were made by means of traditional, direct addressing and strict-behavior programming, related to modeling IoT devices only. Once a typical IoT device is of very limited hardware (e.g., with no screen/keyboard), and the number of different devices at-the-place may be huge, such solution is very inconvenient both for the programmers, and the end-users.

In DIT we propose a new idea to semantically link IoT devices with the real-world objects the devices are related to, such as the doors, windows, lamps, heaters, etc. In our proposal, two basic types of entities are modeled, directly or indirectly via semantic descriptions: IoT devices and their groups, called conglomerates, on the one hand, and real-world objects and their relations to the devices on the other hand. Descriptions of all the modeled entities form a single ontology. The ontology is also responsible for modeling the relations among the devices and the corresponding real-world objects, and, indirectly, the relations among the humans and the objects. The relations are stored in the form of pairs composed of function type, and object type. A function reflects detailed, user-defined action, such as “open”, “close”, “switch on”, etc. Object type reflects a group of real-world objects of certain characteristics, such as “doors”, “lamp”, etc. On the one hand, a pair function-object may be related to a device (or a set of devices), expressing (statically) possible capabilities of this device. On the other hand, such pair may be also linked with an incoming request, expressing requested functionality (dynamically). By comparing, via the ontology, devices’ capabilities with the requests, it is possible to dynamically choose the devices that are able to fulfill the requests. To restrict possible choice, a context is also taken into consideration. Typical contextual parameters are related to geo-location, current environmental parameters at-the-place, such as temperature, lightness, pressure, etc. However, context may be extended as needed, e.g., to reflect number of persons at a given location (this information may be dynamically gathered from specialized IoT devices), hour/date, etc.

For example, a request to “open doors” would be executed in the following way. First, a context is determined, such as geo-location of the caller. Then, requested functionality (“open”) and real-world object type (“doors”) are extracted from the request. Next, according to the SITE ontology, IoT devices are found that are capable of performing the requested tasks and related to real-world objects of given (here: “doors”) type. Activation of the devices makes it possible to fulfill the request, i.e., to open the doors at the current location. Note that direct addressing of both IoT devices and real-world objects is avoided.

A prototype environment has been build called SITE (Semantic Internet-of-Things Environment). With SITE it is possible to (1) express the functionality of IoT devices by pointing out only the real objects these devices are connected to and human-related functionality of these objects, and (2) dynamically choose the devices to serve ad-hoc, dynamically composed requests. SITE may be used in several application areas, including intelligent buildings and places with dynamic and ad-hoc interaction among humans and IoT devices, new generation of museums and exhibitions with intelligent guidance, work-place support at a building place, amenities for tourists, etc. SITE is now intensively tested, also as a student laboratory in DIT. In this laboratory, a modern intelligent building is simulated with dynamic control via direct (e.g., a press of a button), and indirect (e.g., a presence of a human at certain place) requests.



Smart Multi-Device Applications

An issue related with the growing number of interconnected devices forming the Internet of Things (IoT) is to find how users can directly benefit from all these connected devices available around them.

The research goal in this area is to develop a solution for building applications that could be used via any devices available around a user at a particular moment. Currently people use applications developed for specific devices – an application developed for PC or another application for Mac, an application for iPad tablet different than that for Android smartphone. Following the latest trends, a number of separate applications for different TV sets and different car computers will shortly appear. The problem is that there is no single application producer able to create versions of its applications for all possible platforms and devices. One way of addressing this problem is to detach applications from access devices. This approach is used by web applications that allow accessing user’s favorite applications (and user data) in a uniform way on any capable device. However, even with this solution there are still many limitations. Web applications treat different devices merely as different displays and do not take into account their specific capabilities nor the context information they can provide. For example, one can use a web-based TV guide application on a smartphone or on a TV set, but cannot use the smartphone touchpad and gestures to switch channel listings displayed on the TV. Or, consider for example a jogger listening to some music played from the smartphone via earphones. If he/she likes to enjoy the same music after coming back home on home audio system, he/she has to manually search for a specific track and rewind it to play from the point where he/she stopped the music on the smartphone. In future, when all the things will be interconnected via IoT, why not just automatically redirect the music from the smartphone to the home audio system when coming back home and taking off the earphones?

Having all the devices interconnected via IoT, it is possible to share any information among them and to gather external information from numerous sensors available around. This means that one can compose context information and make applications smart (context-aware), maintain application state across devices and use more than one device to interact with an application. The communication technology, various sensors and actuators are already available. The problem is how to build applications that can work in such way and how to make existing applications compatible.

The research conducted in DIT in the field of smart multi-device applications is focusing on the following topics:

  • How to design and implement applications so that they could be used via any subset of devices available around a user at a specific moment;
  • How to make applications context-aware while ensuring privacy and security of user data;
  • How to ensure device ownership and usage rights.



Past experience


Application of Electronic Business Technologies to Supply Chain Management

DIT was involved in application of electronic business technologies and solutions to supply chain management.
After rapid progress in the development of electronic business technologies in the 1990s, nowadays we can see more and more examples of exploiting these technologies to conduct business processes. The possibilities of the Internet – immediate exchange and sharing information among business partners – force companies to reexamine business processes conducted along supply chain, reengineer and optimize them basing on new criteria and parameters that were unavailable before.
DIT research is focused on potential transformations of such business processes. From among potential transformations that are technically possible, DIT is looking for those that are justified from the managerial and economical points of view, taking into account characteristics of local markets. DIT latest results are the quantitative and qualitative models of different variants of supply chain transformations due to electronic business technologies. The obtained theoretical results were practically verified by one of DIT’s business partners – Philips Lighting Poland.

Cooperation with Industry

DIT’s industrial partner within the area of supply chain management and electronic business research was Philips Lighting Poland (PLP) – the world’s biggest manufacturing and logistics center of the Royal Philips Lighting Company – itself the world’s leader in the lighting industry.

The cooperation between DIT and PLP started in 1998. At that time PLP had already perfectly organized and been running internal logistics system but their communication processes with external customers suffered from many shortcomings. PLP addressed a request to DIT to work out new business models which could improve the situation. DIT suggested several possible transformations of the external Philips Lighting Poland’s supply chains, exploiting the potential of electronic business technologies. After thorough analysis, DIT focused on business processes reengineering concerning sale activities, and development and deployment of the Internet e-Order Management System, called e-MAX.

The first instance of the e-MAX system devoted for Polish customers only became operational in the summer of 1999. The current version, called e-MAX-ML (where ML stands for “multilingual”), services customers from the whole region remaining under Philips Lighting Poland management, i.e., 27 countries from Central and Eastern Europe, Middle Asia and Middle East.

The e-MAX system is available for Philips Lighting authorized business partners to let them do business with Philips in a new manner. The e-MAX system permits them not only to place orders, but first and foremost to monitor and negotiate order fulfillment conditions. In the e-MAX approach, all the orders are considered conjointly as a process of continuous negotiations between two parties – the producer (here: Philips Lighting) and the customer (here: Philips Lighting business partner). These negotiations are required regarding availability of products, delivery value and volume, and customer financial conditions. For instance, some products may be available immediately, if they are in stock, or after some time needed to get them from another factory, otherwise. Delivery volume must be optimized due to truck capacity. Delivery value cannot exceed the available customer credit.





Agent Technology for Mobile Assistance

Agent technology was the area of the important research interests of DIT. A software agent is an autonomous program that independently performs given tasks at given place and time according to the orders of its owner. Agent system developed and implemented at DIT is composed of agent server and the database of agents (one agent per user, called “agent owner”). An agent contacts its owner by the use of the SMS and e-mail in a semi-natural language, or WAP/WWW via multi level, pull-down menus, depending on owner’s needs and preferences at given moment and in given place. The system is equipped with multi-level security and safety mechanisms, working both from the outside (access authorization by the use of the phone number and a password, both for the server as a whole and for particular services), and from the inside (access rights to execute given services as well as to use given information coming from these services). Agent key information, as password or PIN, may be ciphered, which raises security level by blocking unauthorized access for other users, system administrators included.

Emerging agent technology for mobile devices, developed at DIT, allows to reverse the way Internet provides its users with information. Nowadays, when a user needs some information, he/she browses the Internet. Agent technology permits to reverse the mode of operation: if something interesting happens in the Internet, the user will be informed about this fact by a software agent via his/her personal mobile communication device. The key question, however, is whether it means something interesting. Note that, first, for everyone it means something different. Second, everyone may change his/her interest very often. Third, for everyone interesting information may come from different sources on the Internet. Finally, for everyone information has to be provided in a different form, because of different communication devices and cultural preferences of a user. Therefore, the characteristics of services provided by the DIT agent server are the following:

  • mass personalization – there may be millions of users registered for a service, each of them provided with personalized information interesting for him/her at a given moment;
  • individual asynchronism – once a piece of information in the Internet is changed and there is at least one service subscriber interested in that information change, a message is sent to his/her personal mobile communication device in a right form;
  • self-management – at any moment, service subscribers may redefine information that are interesting for them directly from their personal mobile communication devices or any PC connected to the Internet without intermediation of a system operator.

There are two new groups of services which may be built basing on the agent technology: personal information services and multi-user communication services. Personal information services may be used in such domains as: banking (alerts about money incomes, intelligently filtered by the agent), sport (notifications about sport events, e.g. goals in football games, but only those that are of interest for the user), health (telemedicine – individual contact of the user with automatic or semi-automatic health care system), culture (notifications about interesting cultural events in the domain that is interesting for the user), shopping (information about promotions of interesting goods), commerce and logistics (supporting mobile human agents transporting and selling goods), etc.

Communication services constitute the second important group of services that may be built basing on the DIT agent server. These services are shared by several users (owners of mobile phones). A group of such users may be spontaneously constituted ad-hoc at any moment. The agent technology offers particular advantages in the case of high fluctuation of the group members, and in the case of differences in information needs of group members. This is a typical situation of groups of students divided into different teams and workgroups. Note that students are widely using mobile devices and are open for innovations. Tourist groups are another example, as well as workgroups of professionals – ‘task forces”, expert teams, etc. It is also possible to build a new spectrum of mass multi user games for the owners of mobile phones.

The agent system has been tested for 50,000 users (generated automatically). During 1000 hours of continuous work, 50,000,000 SMS messages were sent.





Mass Electronic Negotiations

In the area of electronic negotiations DIT was involved in FP5 Project: COSMOS.

One of the main research achievements in the area of electronic negotiations is development of the multi-facet analysis approach to electronic negotiations.

The multi-facet analysis approach to e-negotiations provides a solution to the problem of mass distributed negotiations via Internet, allowing a high number of geographically dispersed negotiators to work on real-life contracts.

In the era of delocalization and globalization of economy, companies need to negotiate at a global scale. The high costs related to face-to-face meetings can be reduced by the use of the Internet as a communication medium. New tools are, however, required to allow contractors, both from multinational enterprises and from SMEs, to negotiate efficiently in this highly concurrent environment, in which the number of negotiators working on a given contract is potentially unlimited. A real challenge in the area of e-negotiations consists in building negotiation support systems that allow a high number of negotiators to work on real-life contracts, i.e. containing both aggregable attributes (e.g. price, quantity, etc.) and non-aggregable attributes (e.g., legal clauses, appendices, quality clauses, etc.).

Two ideas that are the basis of the multi-facet analysis approach to e-negotiation are the following. First, synthetic views of the negotiation process are needed in mass negotiation of real-life contracts, because of the high amount of data. Second, relationships between offers and counter-offers contain information that can be analyzed more easily than clauses contents, because attributes can be non-aggregable and their semantics is not always known.

A prototype e-negotiations support system based on the proposed approach has been developed. It allows a high number of negotiators to work on a given contract and to perform analysis of various facets of the negotiation process in regard to various analysis criteria. The prototype is ready for large-scale experiments.

Current work focus on group dynamics in mass e-negotiations. During mass e-negotiations, many various groups may coexist. Groups evolve: they may split, they may merge, some subgroups may be created, etc. A model has been proposed to provide support for group dynamics in mass e-negotiations. In the proposed model, negotiation processes are structured via negotiation protocols. Negotiation protocols define the way negotiators may act and communicate during the negotiation process inside a given group and according to their role in this group. A prototype e-negotiation support system, named DynG and based on the group dynamics model – has been developed. It allows a high number of negotiators to work in potentially many groups. It provides support for negotiation protocols, allowing each group to be ruled by a different protocol.

Projects in the area of electronic negotiations: