Heading towards a Smarter Internet…

Sydney - 1 May 2003 - The Smart Internet Cooperative Research Centre (or SITCRC) is funding a number of research projects in collaboration with our industry and university partners to ‘solve tomorrow’s Internet problems today’. A key strategy for all of the research is to put the users at the center of our thinking and establish a culture of user centred design. The Taking Users up the Value Chain SITCRC project is exploring the social and behavioural factors that will facilitate many more Internet users in the Australian markets benefiting from a wider range of Internet based communications services. It is also explore how deeper relationships can be built between users and service providers to facilitate higher levels of transactional services. It is also exploring the way in which better relationships can be built between users and service providers to facilitate higher levels of transactional services. The Measuring Customer Response to Radical Future Technologies SITCRC project is developing methods, techniques and software to enable assessment and prediction of user uptake of radical future technologies. The methodologies will provide more precise estimates of degrees of user sensitivity to key attributes of new technologies in ways that measure both native needs and evolutions of those needs.

The future internet will interact with a multitude of network-enabled devices and deliver a plethora of services in a vibrant competitive interactive multimedia environment. In order to achieve this situation, the Programmable Virtual Network research project is developing software tools to enable the Internet itself to be programmable where the service providers will be the ‘programmers’. In this future environment, there will only be a small number of large telecommunications carriers deploying capital-intensive and large-scale infrastructure who provide wholesale products to the many service providers so that they in turn will be able to meet the business and entertainment needs of their business and social customers.

The Smart and Secure Multimedia Content and Delivery SITCRC project is developing tools to deliver ‘smart’ multimedia content in which several resources or ‘digital items’ such as music, video, images and text are packaged together. This smart content will be available for access in all sorts of user environments: from the office, to the home, to the café, but not necessarily through a computer. Today, the Internet only allows the delivery of media files such as music tracks, sound bites, or movies, in the format in which they are uploaded and usually as a single file (such as MP3 for music). Content also now comes in separate files, so if the end user doesn't have the right hardware device, the right software, or enough bandwidth, then they are unable to use or receive the material. Researchers within SITCRC are actively involved in developing and producing software to resolve this problem.

An important outcome of our research into smart content will be that larger content packages, such as whole CDs (rather than single music tracks) will be able to be delivered over the Internet in a single bundle. Security and payment mechanisms for controlling access to and authenticating digital items will also be incorporated into the software so that the producers of the content can receive a payment commensurate with their work. Video files, for example, will be automatically available in the appropriate resolution; high quality for viewing on a high definition video display and highly compressed for viewing via a mobile device. Users will be readily able to switch from one device to the next without interruption to the video. Importantly, the technology we are developing will also make it easier for users to move video, audio, graphics and text around networks and between different file formats.

The Smart Internet will differ from the current ‘dumb’ Internet because in addition to smart content, there will be smart personal ‘software assistant’ (or software agent) that helps you navigate the web and individualise the content according to your personal preferences. It will make sure that the information is tailored to an individual’s needs and present environment. For example, a software agent could be used to regularly search a particular list of web sites looking for any new information (as defined by the user’s preference) that may be of interest. In another situation, the user may want to go on a holiday visiting a given set of cities and places. Having knowledge of the user’s budget constraints, possible departure times, length of stay in each city, desired quality of hotels, and nearby places of interest to visit, the ‘software travel agent’ would search the Web for all the relevant information and provide a detailed report of possible options. The resulting Internet will be less confusing, have simpler access with no Wide World Wait, and therefore will have much wider appeal and relevance in the community.

Through our Digital Rights Management (DRM) SITCRC project, we are also developing software systems and tools that will manage the access rights for multimedia information in a secure and efficient manner. These developments are geared towards providing a software platform to implement systems that will conform to the developing international standards, and will be able to be easily deployed to protect the ownership, and the confidentiality rights of creators, distributors and manage the commercial rights of the owners of the digital content. The DRM system will ensure the authorized use of digital content, and will use a wide range of security technologies including cryptographic algorithms and protocols, watermarking techniques, and tamperproof hardware and software. The SITCRC is developing a DRM system that will protect a single type of right and enable a single user payment for a particular type of content such as the copyright of an image in a web distribution system. The DRM system will also be used to enforce a richer set of rights on a more complex digital object; for example, access to a single image within a photo album with charges paid by a customer.

The business case for DRM is rapidly developing so that DRM systems will soon become a pervasive technology and an integral part of the corporate and commercial world. Security and fair payment of business and multimedia content are the main factors for the current slow uptake of content distribution systems. The SITCRC is also developing models and architectures for rights management to operate in a programmable network environment in which the terminals have varying hardware configurations from a desktop computer to a handheld device or a mobile phone.

The term ‘e-learning’ is often used to describe ‘learning using the Internet’. In this context, a major discussion point concerns the concept of a ‘learning object’. Any efficient and useful e-learning implementation requires a significant investment in the creation of educational content, and in order to make the investment worthwhile, there needs to be a way in which the basic units of content can be readily packaged, transmitted and re-packaged for use in another educational setting. Such a unit of content is referred to as a ‘learning object’, and appropriate international standards which encompass all the important characteristics for the e-learning community are now under development.

However, in the current e-learning standards, a multimedia framework which will not limit the broader range of reusability issues, and a learning framework which will link learning design and digital items is still missing. In our Smart Learning Design Framework project we are developing software that will be able to be used to not only support quality learning designs for all learning sectors, but will also manage intellectual property rights and take advantage of future networks and devices - a priority for Smart Internet CRC research and development.

In the ubiquitous Internet where wireless and fibre optic cable will be everywhere, and the movement of traffic in and out of the wireless component will result in complex and interactive impact on the traffic patterns and utilization of the Internet. The limited bandwidth capability of the wireless component of the network compared with fibre component will also mean that the wireless will need to be managed effectively so as to maximise its value. To date, a lot of effort has been put into wireless resource management to optimize the technical quality of service criteria such as throughput or information transfer rate, utilization of bandwidth or time delay between information source and receiver.

However, economic optimisation has been applied only through simple, static contractual Service Level Agreements and price choices. The SITCRC in its Yield Management for Wireless Networks project is developing pricing strategies for wireless networks by adapting strategies that have been developed in other industries. We are developing an adaptable pricing regime that maximises the economic yield of the wireless network by applying access and usage pricing which reflect both congestion and the consumers' willingness to pay based on technical and economic considerations for multiple applications across multiple networks

In an environment where the Internet is everywhere, the user may attempt to access the Internet via a workstation, mobile phone, a personal digital assistant (or PDA), or some other appliance. The user may also be in the car, at a meeting, or in the street. In this Smart Internet, the location and type of device will place constraints on the mode of interaction between the user and the software personal assistant. The emphasis will therefore be on the portability of information and the mobility of the user. In this situation, the most obvious form of interaction will be by some form of flexible and adaptable speech device, or what we call a ‘conversational agent’.

The SITCRC has produced a prototype platform with facial animation and speech recognition called InCA, or Intelligent Conversational Agent, and has shown how the context is critical for both improving the quality of the experience and the speech recognition rate. In the future Internet, multi-modal interfaces which incorporate various environmental sensors will enable the conversational agent to better inform the speech recognition system of the words and phrases it should expect. As a consequence, the accuracy of the speech recognition, and therefore the usability of the conversational agent will be improved.

Currently, users have to think about the modalities and limitations of the devices they are using and the context in which they are being used. For example, many devices are now capable of text-to-speech as well as having a visual display. It is clear that, for example, when a user is in the car, it would be far more useful for the system to read out and interact aurally with the user rather than using a visual interface. On the other hand, during meetings, a visual medium is usually superior. The Multimodal User Interface SITCRC project will link various input and output modalities to the conversational agent in a natural and transparent manner which will be smart enough to know which modality is more appropriate for the user in the current situation.

We already know from existing research that multimodal inputs can significantly increase the recog