Human-Computer Interaction Books

This book is enlightening for urban studies in the future. Clearly as it suggested the urbanist shall be (able to be) more sensitive to non physical spaces as well as 'conventional' spatial interventions. However it is also quesionable in what way we can catch up with the fast, unpredictable development of internet (along with other h-tech industries which has already gone far beyond our imigination), a review on the developed network is self-evidently far from enough. In reality internet users,at the same moment,urban space users learn cyber-space by clicking, browsing, online-purchasing rather than reading, apparently it is more effective and empirically more understandable. This fact makes theories about cyber-society more or less, inevitbly obsolete and seemingly less neccessary. After all the book is intelligent especially when it explicitly, or implicitly indicates the interaction (possible interations) between urban space and cyber-space.

To someone bewildered by the continuing pace of technological changes, Mitchell offers calming and exciting insights as to where we might all be going. You might well call him a futurist, along the lines of Alvin Toffler and Herman Kahn. Certainly, he writes gracefully, and does not drown you in technobabble.

He describes the rise of a pervasive mobile networking environment around each of us. Encapsulated for the most part in the already ubiquitous cell phone. It is already been noticed by others that cell phone styling is of importance to some users. It bespeaks a fashion sense about themselves that they wish to proclaim to the world. Mitchell suggests that such attitudes will grow, as some manifestations of technology become ever smaller and more closely associated with their users.
A humanising, and not a de-humanising trend.

Too many IT systems have failed because assumptions (often tacit), norms and
values of IT system designers and developers differ from those of their
various business and organizational customers. This has been recognized both
in academia and in industry, and the book's thirteen papers propose
interesting and promising ways of improving this unfortunate state of
affairs. In many cases, the authors of this thought-provoking book do not
deliver answers, but rather recognize and clearly formulate problems, thus
leading to an essential framework for solving these problems. Several papers
provide sincere and often eye-opening assessments of important IT usage
failures. And all authors show interesting and useful models that help to
determine how business and organizational issues -- including political ones
-- ought to be treated in an explicit, rigorous and proactive manner before,
during and after IT system design and development.

E.W. Dijkstra, one of the founders of computing science, noted a long time
ago that when computers first appeared the goal of our programs was to
instruct our machines, but now the goal of our machines should be to execute
our programs. Similarly, the goal of IT systems should be to satisfy
well-defined business needs instead of -- as it often still happens --
directing businesses in their operational, tactical, or strategic work. When
IT provides opportunities as a business enabler, the business (including
social) and IT aspects of an organization are intertwined and thus have to
be
specified and reasoned about explicitly. In all cases, reasoning about IT
systems (which never exist in isolation) and their organizational contexts
should be done using concepts and terms understandable to all stakeholders,
from business decision makers to IT developers. This interesting book shows
various approaches of doing just that.

excellent reference even if old it has a realistic and complete approach to human computer design, including activity. Much better than superficial analysis like heuristic evaluation.

i think this is best book onthat subject ion the 1991.....although some places the explations are not much clear but as the whole concepts/theory given in this book good for students.i

There is no better way to start learning about VR than starting with this book. Ron Wodaski provides great tutorials for using great VR software such as Virtus Walkthrough and Vistapro, both of which are included on the CDs. Also included are chapters dedicated to the big, expensive 3d rendering program, 3D Studio(not included). Possibly the best part about VR Madness '96 is that it comes with 3 CD's loaded with VR apps, utilities, and games.

This book was my first true adventure into making VR a reality on my PC.
It contains a lot of information on gadgets available, concepts of gaming/applications, and the CD's are loaded.

Artificial intelligence (AI) is exploding, both in theory and in applications, as a perusal of the research literature and the availability of commercial products will reveal. The progress has become so rapid that the literature on it has become staggering. This collection of articles is representative of some of the most important developments in the field, with emphasis on theory as well as 'emerging' applications. Readers will probably not be interested in every article in the book, but will instead peruse the ones that are more geared toward their interests or speciality. The section on learning was the one of primary interest to this reviewer, in particular the article on hierarchical hidden Markov models by H. H. Bui, D. Q. Phung, and S. Venkatesh.

As these authors view it, a hierarchical hidden Markov model (HHMM) is a generalization of the ordinary hidden Markov model wherein a hierarchy of hidden states is included. The authors propose a more general HHMM than what has been done in the literature, and one which allows 'sharing of substructures'. This sharing they argue results in more manageable models, and decreases the sample complexity needed for learning. In fact, the complexity of their model is linear in the number of states, making it much more palatable for use in real applications. They illustrate the advantanges of using their version of the HHMM using simulations.

For a particular model depth D, the HHMM consists of a collection of states at each level, with the top level having one state, and the bottom level having D states. There is of course also an observation model consisting of Y observation symbols. One can also speak of the parents and children of a particular state. The parameters of the HHMM give it a definition as a joint probability distribution over a collection of variables that represent its stochastic evolution over time. The authors represent this probability distribution as a dynamic Bayesian network and the HHMM defines a joint probablity distribution over the set of all variables (I.e. the "data") following the factorization of the Bayesian belief network. The conditional probability of a variable given its parents models the state evolution over time, and a state can end only when the state below it has ended. The probability of this occurring is given by a 'termination parameter'. A state will stay the same if it does not end, and if it does end and its parent state stays the same, it will make a transition to a new child-state with the same parent. If it does not end, it is initialized by a new parent state.

The authors then show how to calculate the 'expected sufficient statistics' of the HHMM, where it is assumed as usual that the full set of variables cannot be observed. The algorithm that they use to do this is called 'asymmetric inside-outside', and differs from the usual factorization algorithms used to decompose HMMs. They also tackle the numerical scaling problem that is present in ordinary HMMs, namely the numerical underflow that occurs as the length of the observation sequence increases. The authors use simulations to check that the parameters can be recovered accurately using their approach, and to compare their method with the linear time method. They also give a very short overview of an application of the proposed HHMM dealing with the learning of an HHMM for movement trajectories in a simulated airport environment. They mention an application to human movement but do not elaborate on it in any detail.

Most of the developments in artificial intelligence have been restricted to one domain, such as chess, medical diagnostics, or network event correlation. It is of great interest to be able to design an approach to learning that will allow the machine to reason in more than one domain with little or no change to the structure of its learning algorithms. The article on 'fibring' neural networks by A.S.A Garcez and D.M. Gabbay seems to show promise in this regard, even though the authors' goal is to develop a hybrid neural network that can perform symbolic processing. The allusion to 'fibring' in their approach refers to the ability of single neuron to behave like a complete embedded network according to a particular 'fibring' function. The output of the embedded network depends on the fibring function. The fibring function is not a complicated one in terms of its bare definition. For two neural networks it is merely a function that takes input potentials in one to the weights of the other. A neural network B is then said to be 'embedded' in another neural network A if there is a fibring function from A to B and the output of a neuron in A is given by the output of B. The network composed of A and B is then called a fibring neural network.

The authors give a detailed example of a fibring neural network, emphasizing how several neural networks can be embedded into a single network, and how they can be used to approximate polynomials of arbitrary degree. They also define 'nested' fibred networks as being essentially a series of embedded networks threaded together with a fibring function, and the dynamics of fibred networks is discussed in detail. They then show that fibring neural networks are not only univeral approximators but that they can also approximate any polynomial function to any desired degree of accuracy.

The book provides a good overview if you want to find it all in one place. The authors sound familiar if you attended the conferences or read the proceedings of the AH and hypertext conferences of previous years. Good selection, good reading and a lot of helpful references.

If computing is to achieve its fullest potentail, it is vital that EUC is leveraged to a greater extent than at present. Certainly, the EUC area has matured considerably in the past few years. This maturation is evidenced by this book which makes a vital contribution. It is unique in its focus on a broad range of areas: medical informatics, the contribution of various factors to succesful EUC development of IS, and the role of tools in EUC. It represents the leading thinking by a range of expert researchers and practitioners in this vitally important area.

This product is thicker than i realised. I required the book urgently but this was not in stock and had to be ordered. The book was cheap but i waited 3-4 weeks for it to arrive and the price of the product should have been reduced for this reason.

Despite this, it gives great insight into the application domains and application technologies researchers are using in order to achieve Affective computing within Gaming and Learning environment. I would have liked to have seen more information in relation to the background and theoretical aspects and affective computing relationship with HCI since information within this area is lacking. However, it is a good book.

It is always fascinating to look back at OS/2 to see what a miserable sub-set of IBM's product Windows represents.

Remember Conway's Game of Life? Surely you must, if you are interested in this book. The Game has been around since the 70s. The editors have cultivated recent research papers that demonstrate how far the field has advanced. Reinforced by some pretty colour plates that depict artificial entities [dare we call them living?] in some surroundings. These include the modelling of bee flights through a garden, and simulated trajectories of a group of bacteria.

Nor is the Game of Life ignored. One plate shows it in three dimensions. The Game is played in 2 dimensions, with time as the third dimension. An obvious choice that gives interesting trajectories of the cells.

The narrative adds to the illustrations. By describing a variety of computer simulations [worlds?]. Where the experimenter can tweak many parameters, and watch her world unfold. Some worlds are impressively rich in complexity of observed behaviours.

The only drawback in the book is its skimpy index. A mere two pages. It should have been more detailed.