Visualization is the key methodology that gives the researcher insight into data that may be generated from various sources such as computational simulations and scientific experiments. A recent trend has been towards the use of VR technology for data visualization to give the user a realistic insight into the data.

Teleimmersion is the amalgamation of data mining and significant computation with collaborative VR. It allows multiple networked users to participate in a shared virtual environment. The collaborators can talk to each other and can see each other in the environment. Teleimmersion augments the data visualization and analysis process to produce a new genre of applications.

With advances in the fields of computational science and engineering we now have faster computers that generate data, which is in the range of a few hundred megabytes to several terabytes. Large data visualization poses a new challenge to the visualization community, as most of the existing systems are not capable of visualizing vast amounts of data.

Massive data sets and collaborative visualization add a new dimension to ongoing research on visualization in virtual environments. Collaborative extensions have been added to existing non-VR systems. A number of dedicated single user systems allow the user to visualize large datasets in virtual environments. Most of these systems are application specific and cannot be extended. Even though these systems cater to a diverse set of application domains, some trends and patterns in their approach to visualize data are discernable. These design features can be reused in future applications.

The main contribution of this thesis is in proposing the Teleimmersive Data Explorer (TIDE): a general architecture that blends collaboration with visualization, which can be used by application developers for rapidly building teleimmersive applications for large data visualization. A basis of the TIDE architecture has been implemented.