|
![](images/pix01.gif) |
|
|
|
Application of Neural Network and Its Extension of Derivative to Problems of Visual Realism on Web Pages |
|
PP: 1993-2000 |
|
Author(s) |
|
Lan-Ting Wang,
Chang-Franw Lee,
|
|
Abstract |
|
This study utilizes information sciences (i.e., neural networks and extensions) to reduce the burden for investigation of social
sciences. Without loss of generality, the visual realism of pictures is utilized as the example to illustrate the proposed research flow
chart. Pictures of web pages are important media for conveying meanings of information systems. Visual realism of pictures means the
degree of similarity between pictures and real objects, and is an important topic in designing contents of information systems. In this
paper, the neural network and its extension of derivative are applied to participants’ preference for visual realism on web pages. This
study includes three parts. In the first part, the participants’ preference for visual realism on web pages is experimentally investigated.
The investigation results are recorded, analyzed and discussed. In the second part, we utilize the neural network to model and predict
the investigation results of the first part. With the help of neural networks, one can model and predict the full investigated results from
only part of the investigated information. In the third part, we develop a neural-network extension of derivative to predict the increasing
or decreasing trends of investigation data. With the help of neural-network extension of derivative, the trends of investigation data can
be predicted without plotting the overall data curves. It should be noted that the neural network and its extension of derivative share
the same training procedures. No additional training work is required in the neural-network extension of derivative. The use of neural
network and its extension of derivative will greatly reduce the investigation efforts and one can obtain almost the same results as full
investigation. This study will be helpful in understanding, modeling and predicting participants’ preference for contents of information
systems. Although only specific examples are illustrated in this study, the proposed research flow chart can be applied to many other
problems of questionnaire investigation in social sciences. |
|
|
![](images/pix01.gif) |
|
|