Learning in a Highly Interactive, Low-Resolution Virtual Environment; The Tomb of Lady-Hao

We are interested in users learn, classify and recall spatial and attribute information about objects in a virtual environment which is highly interactive, but built with minimal detail. The Hao Tomb application represents spatial relationships between objects with two-dimensional images (billboards) which are embedded in a three-dimensional model. Information about these objects, (i.e. materials and function) are represented through a visual query mechanism. Given the task of learning certain attributes of the objects found in the tomb and their spatial layout, test subjects using the Hao Tomb application did better than subjects given high-quality printed information. Given the small number of subjects, these results are preliminary. They are part of the lessons learned in a larger pilot study, which has paved the way for a more in-depth study. INTRODUCTION & BACKGROUND The purpose of the Hao Tomb application itself is to make a body of knowledge more accessible through the use of an interactive, spatial interface. Archaeologists studying documentation off a dig need to know what items were found and where they were located. The application presents a subset of key objects found in this particular tomb and allows the user to perform visual queries on their primary attributes. It is our belief that the objects' orientation in three-space are best learned by looking at a three dimensional model. Also, the ability to perform interactive visual queries on the attributes of the objects should enhance learning of those. Neither idea is new. [][][][][] Our primary innovation is use billboards, simple onedimensional icons that always face the user. While they have been long used in virtual environments, hence a specific name for them, it has usually been to represent a distant object, or out of necessity or simply to enhance performance. Most developers, and especially those of archaeological simulations, prefer to render all objects in a scene in a full three-dimensions, with as much resolution as possible. While this is desirable for many reasons, it requires a lot of computing power and developer time. Instead the time and cycles saved by using simply billboards could be used for greater interactivity and the make the application more widely available. The goal of the Hao Tomb project is to prove that this is a desirable strategy. Source Material The virtual model represents objects found in the Bronze Age tomb of Lady Hao, a principal wife of Wu Ding, a famous Emperor of Shang dynasty China. It is the only intact royal Shang tomb excavated (in 1985) using modern archeological methods. Spatial relationships are particularly important in it because the objects were found in several distinct layers, which scholars believe corresponded to phases the Shang burial ritual. Dr. Linduff's contention [] is that Lady Hao was not only a warrior, and probably a general, but that she was probably not Chinese. Both assertions are controversial and current among scholars, and the 23 objects represented in the application were chosen to make the point. THE HAO-TOMB APPLICATION The Hao Tomb application is a web page with a large VRML component. All of it is freely available at "http://www.sis.pitt.edu/~goshen/Net-Hao/". The initial page contains two main panels, where the right contains the VRML model of the tomb. (Fig 1) The actual “tomb” was really a large pit, filled in with eight layers of dirt and roughly 1500 objects placed at each level. This model only shows three layers and 23 objects. The layers appear as “floors” of a translucent mesh. Billboards represent the items on each level— small images of the items placed on a polygon that always faces the viewer. At present, all billboards are the same size, regardless of the actual sizes of the objects themselves. Each icon is placed in the approximate location where the object was found in the tomb. The mouse appears as a small arrow on the screen. When it is over an object’s billboard, it turns into a small circular sunburst figure. At that point, the user can click on it, which causes two things to happen: At the bottom of this window are the standard navigation controls for he COSMO browser. Full instructions are available at: "http://cosmosoftware.com/support/ player/docs_21/frames-help.html". Controls are set to “examiner” mode, which allows the user to rotate the entire tomb to any angle and to zoom in on any part of the model. • In the left-hand panel of the page, a larger image of the item and a short paragraph of information about