Year I

• The Alexandria prototype will be implemented based on the preliminary user requirements and design specifications. Organizations (e.g., public, academic, and special libraries associations; AAUP) will be alerted about Alexandria's goal and time frame, and their evaluations will be requested. We will compile a questionnaire, send out for comment, and use this information as input to the design process. The prototype will be completed at the end of the first year.

• Investigation of wavelet transforms and subband decomposition techniques for efficient storage and retrieval of images.

• Investigation of the functionality of: the user interface; the procedures required to extend khoros procedures; and the hierarchical data structure that supports browse on graph-based datasets.

• Initiate the conceptual design of: the text-based query language (extended SQL or MDBL); the visual query language and in particular the language GQBE; and the browse system for raster sets based on wavelet transformation.

• Investigating the suitability of image features detected by nonlinear filters for indexing and development of appropriate quadratic nonlinear filters for enhancement and restoration of satellite images. Investigating low level image features such as texture, color, and contour information in encoding map and satellite data for indexing. Investigating the use of contour based image registration algorithms for registering maps with images as well as image to image registration.

• Evaluating various spatial index methods for image data and designing a multidimensional index structure based on the extracted image features.

• Assess the feasibility of broadband transmission systems and determine the tradeoffs between speed, access time and number of users. Evaluate the tradeoff between computational effort, storage and bandwidth of the delivery system. Determine the threshold at which compression becomes advantageous. Evaluate existing network authentication mechanisms and protocols.

• Investigate performance issues, identify parallelism in each component and between components of the proposed system, and study the feasibility of parallelization.

• Measure user workload requirements, determine size of datasets, and develop a model for catalog accesses Incorporate workload and dataset/catalogue models into a system model in order to analyze capacities and throughput. Use system model to identify bottlenecks in processing power, storage capacity/bandwidth, network latency/bandwidth, and algorithm performance.

• Review current information practices and the existing literature, expose reasonable alternative information policies and actions, predict short-term and long-term ramifications of those alternatives, and articulate arguments for and against competing alternatives. Prepare comparative analysis survey/case study instruments.