[ Home | Search | Contact Us | Index ]
[ Sutradhara | International Dialogue Unit | Multimedia Development ]
|
Home > Cultural Informatics - Multimedia Development > Multimedia for Humanities >
( Download PDF file) |
|
A Model Design for Computer based Cognition Support Systems
Sanjay Goel, IGNCA |
|
1.0 Introduction Different life forms have been gifted with a different mix of powers to sense the physical world and interact amongst themselves. Intra-species Interaction among all life forms is psycho-physiologically essential and natural. Most primitive forms like amoeba can sense the environment only by physical touch, whereas advanced life forms have more senses. Man has been gifted with five senses to register some characteristics of the physical world. Though other animals specially mammals and birds also have five senses, the resolution, accuracy , range and bandwidth of each sense differs in different species. Because of the great gift of an advanced brain to man, his psychological needs for interaction and communication far exceed the physiological reasons. A very significant amount of human effort is targeted towards pursuit of knowledge. Pedagogical communication is the most important form of formal communication. Fast growing corpus of human knowledge offers interesting communication design challenges to academicians and designers. While animals can communicate with the limited sounds, rich spoken languages are nature’s unique gift to man to facilitate communication. Since the early days, man has been generally communicating through oral and/or visual expressions. Natya Sastra, the ancient Indian text on performing arts, had limited the sensory scope of artistic expressions to eye and ear only. Empowered by his unique ability to make tools, it was most natural for man to develop tools to record, store, transmit, render and disseminate these expressions. These tools allowed him to communicate across the barriers of physical space and physical time. Though research to develop the tools to support other senses is also in progress, the tools for audio/visual support are sufficiently mature and continue to draw the attention of the largest number of developers for further improvements. The word "Multimedia" has become popular in the context of concurrent presentation of multiple audio/visual elements to formulate and communicate richer expressions. Communication "Content" may include data, information, creative expressions or analytical interpretations in different media. Data refers to factual representation of reality, usually it is captured mechanically and may not be much useful in its raw form. Information is mechanically filtered out of data to provide a focussed answer to a specific query. Analytical interpretations result out of scientific analysis of reality. Some of these discoveries can be proved to be factually correct and may get accepted as data. Creative expressions reveal intellectual and creative faculties of human mind, and may be partially or fully inspired from the reality. They are highly subjective and allow a non-disturbing co-existence of contradictions. The notion of consistency and completeness can not be rigidly applied to these expressions. Designs are a form of creative expressions with more or absolute emphasis on consistency and completeness, and are grounded in deep analytical understanding of the reality. Analytical interpretations and creative expressions result out of human intelligence and can not be mechanically inferred from the data and/or information. Researchers of Cognitive Science and Artificial Intelligence are trying to model and partially simulate these human capabilities. Communication technology has been history’s driving force[ ]. Man took almost 50,000 years to start the development of scripts in ca. 3500 BC. Well structured scripts based on proper alphabets were developed around 1000 BC. Because of the power of the written word, the next significant progression was made in just 2500 years when printing technology became available around 1500 AD. The 19th Century empowered us with machine aided recording and communication of audio/visual details. Electronics started another revolution in the late 19th Century and Human expression could be captured and disseminated electronically. Different technologies were developed for "soft printing" of different kind of expressions. Electronic technology has now reached a maturity level that empowers us to handle different expressions almost in a uniform manner. All these advancements have helped in democratisation of the access to learning material and an increase in the general education levels. To support human communication , the electronics industry has made significant progress in Telecommunication, Consumer & Entertainment Electronics and Computers. Technology has reached a certain level of maturity in each of these areas. Since the very beginning there have been occasional exchanges among these areas; however, it is only recently that an integrated approach for further development has started. The Publishing industry has undergone a revolution in the last twenty years because of the integration of the Printing Press with Computers. Until recently computers did not support high capacity, portable content storage devices. Because of this limitation, computers could not be used as communication and educational medium. Their usage remained confined to data processing, information systems, physical process modelling and control applications. With the increased processing power and storage capacity, a new communication medium has become available. Computer based "Interactive Multi-media" is now facilitating the simulation, extension and integration of various communication technologies to facilitate multi-sensory communication and learning. Technologies are not mere exterior aids but also interior transformations of consciousness[ ]. The form and the structure of the discourse has been influenced by the affordances of different tools. All communication technologies have their strengths and weaknesses that are gradually realised after some years of initial experimentation with the technology as a communication medium. It took several decades of book printing for a useful and generally acceptable book form to emerge out of the tradition of costly medieval manuscripts[ ]. For several years, motion picture directors continued to use the new medium as theatre, and it took quite some time for producers to stop using television as glorified radio. Interactive multi-media technology also will soon be able to create a niche place for itself out of the legacy of the earlier media. Depending upon the level of activity by communicators, communication can be broadly classified into the following categories[ ]:
Here, the phrase "passive" does not imply completely inactive receiver; it only reflects a relatively lower activity by them. Unlike the audience of an entertainment programme, a learner actively participates in a symmetric duplex communication with the teacher, peers and pedagogical resources. Learner’s activity is the most significant contributor towards effective learning. Several tools are available to support these activities and reduce the load on the learner and enhance the learning. Computers have the potential to amplify the extent of this support. Computers offer interesting novel possibilities of interactivity and of integration of data, information, analytical interpretations and creative expressions. The issues being addressed by researchers and practitioners engaged in the area of Interactive Multimedia can be broadly abstracted as follows:
A lot of work has been carried out by computer scientists to address the first two issues and sufficient core technology is already available to the designers and developers of Multimedia Communication Systems. Several techniques have been developed for Text Mining, Speech analysis and Image analysis. Rapid technological advancements are empowering the designers with even better data encoding, retrieval, rendering, synthesis and analysis tools. Technology assisted content analysis in certain domains like art history has not yet been much experimented. Though several groups are developing pedagogical multimedia systems, adequate work has not yet been reported to set the detailed design and process guidelines. Several authors have focussed on different aspects, a detailed design model has not yet been reported. An attempt has been made to propose a detailed design model for pedagogical multimedia systems in this paper. 1. Computer based Cognition Support system:Amongst other social activities, education ranks very high as one of the main beneficiaries of advancements in communication technologies. But unfortunately, it has not been the first to take advantage of some of the key advancements. It is the entertainment industry which has made the largest investments in applying the upcoming communication technologies. Though cognition is an internal and subjective process, it gets greatly effected by the way the study material is made available to the learner. Technology, if appropriately applied, can support cognition and enhance the learning. The design requirements of educational software using any of these technologies are not the same as that of an entertainment software. While media experts can exploit the communication technologies for developing entertainment and informative software, the educationists themselves must play the main role in designing and developing the software for pedagogical purposes. While educationists are playing the central role in book publication, they need to closely collaborate with media and technology experts for using other communication technologies. Multidisciplinary nature of developing computer based educational software makes the cost of such a development quite unaffordable by small groups. Interactive multimedia demands specialised technical and design skills, these projects are generally being planned and managed by multi-media experts. Educationists themselves are acting as resource persons with the core responsibilty being with multimedia designers. The success of this technology as an accepted and effective medium of education will ultimately depend upon the level of confidence with which educationists will be able to conceive, plan and manage their projects. It is expected that much of the design effort will be saved with the proposed model and educationists will be able to specify system features more explicitly. With this model, the educationists can take up the core responsibility for content creation and multimedia experts can focus on their task of digital realisation. Learning as an activity starts with exposure and progresses through pedagogical resources with the help of cognitive processes [ ]. Pedagogical resources include study material and external as well as internal cognitive tools. Study material comprises of multiple forms of primary and secondary "Content". Primary content includes the copies of original object under study and Secondary content is the scholarly discourse/ interpretation/ commentary of the primary. In general, only fortunate few learners have been able to have access to primary content. High duplication costs of the non-textual study material has been a major deterrent against democratised learning. This problem can now be solved by the digital technology because rapidly reducing duplication cost of digital content is media independent. It is now feasible to provide instant access to large amount of duplicated primary content to much larger population in a very cost effective manner. Rapidly decreasing costs of multi-media ready computers will soon take them to almost all middle class families of most of the countries. Educational software will take a major share of computer software on such home computers. Because of the powerful and fast-growing technological potential and lack of enough maturity in the area, there is a lack of suitable models for applying interactive multimedia technology to develop educational software. As happens with any new technology, people from varied backgrounds have been trying to develop Computer based educational software with different kinds of approaches. Some use Computer as an extension of Video, some extend the databases to include multimedia objects, while some extend the electronic text to offer a hyper-linked network of visual resources depending on their background primarily being in Video, Information sciences or in Data Engineering. Singularly all these approaches fail to exploit the full potential of the computer as a general purpose Simulator. Initially new materials and media are generally used as a mere substitute for the existing technologies. However, new technologies create a genuine place and social need for themselves only after novel features offered by them are taken advantage of by the Designers. Computer aided Instruction (CAI), Computer aided Learning (CAL) and Computer based Teaching (CBT) are three popular words amongst educationists. Systems developed under these categories do not offer sufficient study material, and also do not seek to support various cognitive processes. The following Notes are a summary of the author’s research towards developing a model design for Computer based Cognition Support Systems. It has gradually evolved over the last three years and is a result of the author’s involvement in the design of several multimedia projects at IGNCA on a variety of interesting subjects in collaboration with some well known art historians, archaeologists, anthropologists, epigraphists and linguists. It is hoped that the proposed model’s applicability is not limited to these domain of Arts only, and it is generic enough to be applied to other disciplines as well. However, the author has not yet had the opportunity to apply the model to other disciplines to confirm the claim. The recently developed and quickly expanding field of Cognitive psychology throws a very useful light on the learning process. A closer look into the process opens up interesting opportunities to the designers of Interactive Multimedia Educational Software. Learning is multistage process and results into formation of higher level mental constructs known as Cognitive Maps. Oriented learner gradually makes a transition from an intelligent but mostly passive receiver to an active information collector and finally to a researcher. Different kind of cognitive activities are performed during different stages. Learning results into mental encoding of knowledge which is internally represented by multidimensional higher level mental constructs known as Cognitive Maps. Faithful external representation of these internal cognitive maps results into efficient teaching. A novel generic knowledge representation scheme to externalise the Cognitive Maps has been formulated and applied to all the ongoing projects by our group. Digital technology is helping us to redefine the notion of the book from a static and linear collection of limited visual content to a dynamic and non-linear corpus of large body of multimedia content. The proposed model has evolved out of the desire to support the learning process at all stages and extend the ‘Book’ paradigm. Computers ability to store large volumes of instantly available data, to represent any structure or behaviour, and, to integrate multiple elements are three underlying strengths on which this model is based upon. It tries to harness the power of interactive multimedia by offering extensive study material and pedagogical tools. Special attention has been paid to enhance the interactivity. Effective mechanisms have been proposed to facilitate uniform and quick exploration, rendering and analysis of large digital corpus of primary, secondary and tertiary content. In short, the model tries to free the ‘Book’ from the constraints of the paper. While there is a lot of activity going on in the development of multimedia presentations, this model can be used to develop serious study aids on any subject. According to Hisenberg, famous physicist of modern times, the synergy between thinkers from seemingly different disciplines has helped in giving major turns to civilisation. Absolutely fresh and new ideas emerge during such multi-disciplinary collaborations. Multimedia development is one such multi-disciplinary activity. Successful educational multimedia projects will emerge as a result of very close collaboration between subject experts, educationists, computer scientists and media designers. This synergy between Computer Science and Humanities is opening up new research methodologies and interdisciplinary research problems for scientists, and as a consequence, is producing far more comfortable, enjoyable and holistic learning for the learners. If the concerned collaborators have sufficient skills to play more than one of the above mentioned roles, the synergy is going to be even more fruitful. While the collaboration between subject experts, educationists and media designers is an established practice, the involvement of computer scientists becomes essential because of the immature level of so called easy to use multi-media authoring tools and proper design models. As these tools gain the maturity to really harness the actual potential of the underlying technology and some projects set the example by offering powerful content exploration tools, computer scientists’ role will become more focused on the development of the development tools rather than the individual multimedia projects. At the moment the responsibility to educate the subject experts and educationists about potential possibilities largely lies with the computer scientists. Soon the subject experts and educationists will start pushing the agenda for the computer science community by demanding more features in such tools. Creative subject experts have played a very significant role towards the development of the proposed model. An attempt has been made to develop reusable generic design specifications for developing Interactive Multimedia Projects. Some of the proposed features have already been implemented by our group . More features are under development. It is hoped that much of the design activity for individual future projects can be cut short by selecting features out of the proposed set. However, creativity never stops and the model will continue to evolve to reach higher goals. The following text presents achievable generic design goals for serious designers. Issues related to Corpus configuration and Media component selection are discussed in ‘section 2’. ‘Section 3’ briefly introduces the complexities of learning process and provide the theoretical foundation for proposed interactivity features discussed in ‘section 4’. Finally, ‘section 5’ focuses on the Quality related issues. 2.0 Corpus Configuration: According to Marshall Mcluhan, a well known media scholar of 1960s, a given technological medium itself determines or in large part influences significantly the message that it will convey. The primary tool for communication is the symbol, and man has been termed as a symbol using animal[ ]. Computers started as storage, processing and retrieval device for numeric symbols and soon were able to handle textual data. With the help of digitisation devices, Pictures, Audio, Video and other forms of expressions can also now be dealt with, although computers still are most efficient with alphanumeric representation. Digital representations of other expressions don’t yet allow a very easily implementable and reliable content based retrieval unless the content is supplemented with textual codification; and also they consume much more storage space per minute of information. Appropriate data-encoding and compression techniques should be used to digitise the content corpus. With careful planning, either of the following data elements can be stored on one CD-ROM, today’s removable data storage medium:
The physical data storage capacity of the upcoming DVD is eight times the capacity of the CD-ROM and the recent trends in multimedia data encoding research promise a much higher density of data per Mega Byte of storage space. Hence, it will soon be possible to store much larger corpus on easily portable disks. The quality of audio-visual elements is also improving because of the increased data density and newer algorithms. Further, the Internet offers the possibility of boundless corpus to be offered to the interested learners. On an average experienced readers read approximately 20 pages in an hour. Hence, it can be seen that 2 Lakh pages of printed text provide a study material for approximately 10,000 hrs. Surely, except for encyclopaedic systems, this amount of study material is not very desirable to be fitted in one CD-ROM. Hence most of the space is usually allocated for storing other media components, e.g. Pictures, Audio and Video. Though the educationists have experimented and utilised the upcoming communication technologies, it is interesting to have a look at the media choice pattern of the learners. Books still remain the most popular medium for self learning i.e. "Swadhyaya". Though available for the last few decades at an affordable cost, Video has not become a very popular educational technology for focused learning. The same is the case with the Audio Cassettes. This is not to be confused with aired Audio- Visual programs through radio and television. While these audio-visual programs are very effective for mass education through broadcasting or occasional group viewing , it is the printed word that has remained the primary tool for personal learning. In the Study Rooms of most of the learners, books and photo albums continue to outnumber the audio and video cassettes. Hence in order to make it a really successful pedagogical medium, the interactive multimedia is recommended to be used as a meta-medium for the development of learning aids based on a design model which is primarily an extension of the book paradigm. In short, it should include the features of currently popular educational technologies and the novel content processing and searching possibilities being offered by the computers. Many designers coming from a video background tend to use a lot of video in the CD-ROM Multimedia systems and very often come across space limitations. A designer should try to arrive at a proper balance of media components for optimum utilisation of the today’s technological possibilities. Properly laid down design objectives help in arriving at right decision. As already stated, video consumes a lot of space and hence it should be used very selectively. Ideally it should be used to demonstrate kinetic action. A well conceived mix of still pictures, music and oral commentary often serve the purpose equally well. However, selective usage of small video clippings greatly enhances the learning possibilities. The uniformity of underlying digital encoding principle offers the unique possibility to mix and integrate different media components at multiple levels. Respective strengths of carefully selected primary medium need to be intensified and augmented by other media elements. Primary medium and complete media configuration for individual discourse propositions can be differently designed. However, attention must be paid to consistency and fluidity aspect of communication. Ability to easily mix varied media elements poses a challenge for communication designers to keep the individual media streams spatio-temporally synchronised at lexical, syntactical as well as semantic level. An under-estimate of the effort required to ensure this causes delays in the projects and often a compromise is made to rush and declare the completion of the project. Even a small error is capable of causing an absolute damage to the message. Hence the designers need to be very careful on this issue since the beginning. Formal modeling and representation schemes must be followed in order to design, validate and verify the synchronisation of media streams. 3.0 Learning Learning is a multistage process. It is highly subjective. One has to be a disciplined learner in order to abstract useful learning. Education and Research in Humanities mostly involves Qualitative Analysis and occasionally uses Quantitative techniques. An interested and oriented learners passes through various stages. As per the ancient Indian tradition Knowledge Acquisition progresses through the stages of Adhyayan (Learning), Bodh (Understanding), Aacharan (Use) and Prachaaran (Exposition). Recently some attempts have been made to model the process. Grounding Theory proposed in 1980s is an interesting attempt to understand the process of Qualitative Analysis. Following stages are proposed by this model:
An almost similar model has been considered by the author for designing the computer based cognition support systems. It comprises of four stages. It is important to note that a definite sequence does not separate the above stages and the learner tends to move back and forth between these stages. These stages are as follows: 1. Experience Accidental or deliberate experiences offer the opportunities to start rewarding journey of oriented learning. First hand experience is caused in the moments of direct encounter with multi-layered and multi-dimensional reality. Discourses, Museums and Exhibitions try to reconstruct the reality in a focused but limited manner and offer second hand experiences. Different reconstructs need to be presented to different audience depending on their socio-cultural-lingual-academic background. Most effective reconstructs have following features:
This phase does not necessarily require high level of active involvement by the audience and the cognitive load is very low as compared to further stages in the learning process 2. Labelling Throughout the process of learning, individual meaning components and configurations are labelled. Conceptual labels are placed on discrete objects, happenings, events and other instances. These are classified into higher order, more abstract Categories. Attempts are made to label the following conceptual entities:
Labels for some key concepts are presented during Discourse but a detailed labelling is explored by the learner during Analysis. After detailed Analysis new labels are assigned by the researchers out of which some may be accepted by fellow researchers and the society at large. 3. Analysis Only oriented and disciplined learners get interested to carry out a detailed Analysis. Syllogistic Reasoning is carried out by the learner to construct a personal piece-wise sequence by gleaning and picking out individual details out of the study material in linear, non linear and associative manner. High level of learner’s involvement is essential for successful progression through this phase. Well structured study material increases the learners’ freedom to access the content as per their requirement. Searchability, flexibility and even the utility of information largely depends upon its structure. Sensitive learners are also interested in Dimensionalising the properties of different Categories and like to see the Dimensional Profile for each occurrence of a category. 4. Synthesis The learner gets an aha! experience, known as Gestalt Recognition in the cognitive moment when all pieces of the puzzle come together. The "Whole" which is different from the sum of the parts suddenly becomes visible. An experienced learner creates personal Cognitive Maps to unify the total context of the knowledge. Even contradictory experiences can get assimilated without impairing the overall unity. Cognitive Maps are synthetic, higher order meaning configurations and are capable of very efficiently encoding the knowledge. These are generally long lasting and easily modifiable. It is a challenge for educationists to defuzzify the given domains and help the learners to construct their own Cognitive Maps. The proposed model also includes a novel generic knowledge representation scheme to externalise the Cognitive Maps of experienced scholars on any given domain. 4.0 Interactivity Goals The first generation of technology gave us devices that were primarily extending our limbs to carry out menial tasks easily, rapidly and from a distance. The second generation’s focus has also been on supporting the senses. It is the third technological generation which is allowing us to extend our own mental capabilities through computers. The Computer is basically a general purpose representation machine i.e. Simulator. Theoretically, it can be programmed to simulate any defined and analysed Process and Environment. Human creativity and Technological possibilities have gradually led to the deployment of computers in almost any walk of life. Gradual advancements are very rapidly expanding the technological possibilities. It is left to human creativity to use these possibilities to the fullest and use the technology in the most meaningful way. Study Material includes all forms of Primary as well as Secondary Content. An active learner applies various external as well as internal Cognitive Tools and Processes and explores the study material. A computer aided Cognition Support System should try to simulate and extend these tools and processes as well as offer an integrated corpus of study material. Following are the proposed simulation goals : A) Study Material At the General Conference of UNESCO, 1964 a ‘Book’ was defined as a non periodical printed publication of at least 49 pages, exclusive of cover pages[ ]. Publications of lesser pages were put in the category of ‘Pamphlets’. An ‘Electronic Book’ is expected to offer much more content as electronic encoding facilitates very compact storage. Though no standard has yet been defined, approximately 1000 printed pages worth of content is proposed to be the minimum for an ‘Electronic Book’. Any thing substantially lesser than this will fall under the category of ‘Electronic Pamphlet’. If required, the constraint of non-periodicity can also be eliminated because of the advancements in computer communication. The CD-ROM content can act as the ‘Base Content’ and subsequent updates can be accessed via internet through the same user interface. The ‘Base Content’ could also be updated at a slower pace and new versions can be released periodicity . Different disciplines give different amount of emphasis to different kind of study material. Generally, learners use following type of study material:
B) Cognitive Tools
C) Cognitive Processes
Inter-disciplinary and Holistic studies demand contextual access to visual, oral and textual sources in an integrated manner. Complex knowledge system of Humanities offer multiple hierarchies of closely inter-linked multi-dimensional categories. An active learner makes an intra-dimensional as well as inter-dimensional progression through these categories. Access to related sub-network is made in the context of the chosen categories. In this process new relationships and categories are discovered by researchers. Based on their prior experience and learning styles, the learners tend to choose one kind of resources as their grounding study material, while the other resources are used to expand the context. Using the flexibility of computerised databases, it is possible to empower the user to choose any kind of resource material for grounding the studies of other kind of resources. Flexibility of Mechanised Content Access can only be harnessed by supplementing the Content with a rigorous Content Structure. These structural elements represent essential knowledge components and convert a collection of data items into a Content Corpus or Content Base. An optimised parent-child causal model often results into an over-simplified representation. Further, it is a challenge for educationists to defuzzify the given domains and help the learners to construct their own Cognitive Maps. Subject Experts and System Analysts have to very intensely collaborate to formulate and represent these knowledge structures. Formulation of these structures and representations is the most crucial phase in the process of designing a Computer based Cognition Support System. Success of the project largely depends on the intensity of this collaboration. Hence Designers are advised to specifically focus and be ready to spend substantial intellectual effort in Content Structuring. The computer relies upon these structures to execute various search related tasks and relieves the learner from lot of mechanical activities. This formalisation also helps the subject experts to discover the incompleteness in the data and creates opportunities to complete the same. The proposed model also includes a novel generic knowledge representation scheme to externalise the Cognitive Maps of experienced scholars on any given domain. As has been discussed in the last section, different stages of learning process demand different kind of support from the learning aids. A complete Cognition Support System should offer support features for all these stages. Hence, the user should be allowed to interact with the system in the following modes : a) Discourse mode (Story Telling) b) Exploration mode (In depth Study) c) Abstraction mode (Internalisation) A learner model inside the computer can be used to recommend aspects that might have been completely missed or not adequately examined by the learner. An optional feature of On-line Continuous Evaluation can be included to offer feedback to the students. [ Next Page ] Contd... |
|
[ Home | Search | Contact Us | Index ] |
[ Sutradhara | International Dialogue Unit | Multimedia Development ] |
Copyright IGNCA© 1999