1 Overview of DANA
DANA is being created through development and implementation of cross-platform, open standards that will facilitate interoperability and information exchange in a distributed network environment. The intent is for the distributed network to connect remote systems via the Internet. DANA will constitute a federation of digital databases located at institutions across the USA and around the globe. Through use of a single, Java-based application, Java Web Start (Sun Microsystems, Inc.1995-2002a), DANA will link researchers, educators, students and the general public to digital databases that include realistic, accurate, three-dimensional (3D) visual representations of artifacts, fossils and other objects, along with two-dimensional (2D) digitized documents (e.g. maps, plan views, excavation profiles, field notes, manuscripts, charts, graphs, and photographs) and associated content.
This online network of anthropology resources will allow reliable "anytime, anywhere" access to content and services. By enabling dynamic use of digital representations, virtual tools, and an array of data supplied by contributing content providers, DANA will be a valuable resource for education and research. It can be accessed and used at all educational levels, from kindergarten, through elementary, secondary, undergraduate, graduate, postgraduate and professional, to members of the general public engaged in lifelong learning. Users can take advantage of the DANA system for formal research or simply to satisfy curiosity. The purpose of DANA is to provide a unique and innovative set of products and services for advancement of cultural heritage management, scholarly research and education in anthropology and related disciplines.
Our current operational prototype for the DANA system works with the Oracle 8i relational database management system (RDBMS) and PostgreSQL. To maximize the number of databases with which DANA will work, the system uses generic SQL statements for table creation, data insertion and querying. We anticipate that it will work with IBM's DB2, Microsoft SQL Server, and other fully functional RDBMSs with Java database connectivity (JDBC) data access API drivers (Sun Microsystems, Inc.1995-2002b). JDBC provides a layer of abstraction to database connections that allows the system to connect to different database software through the same interface.
To aid in the interchange of anthropological data, we plan to define an Anthropology Markup Language (AnthML) created from an XML Schema specification. We will also create a glossary and a thesaurus used in AnthML. Anyone interested in participating in the development of AnthML is encouraged to contact the ATL.
The project team has developed a prototype Data Entry tool for qualified professionals (e.g. collaborators) to add digital content to the databases, either as additional data for inclusion in the existing system or as new data sets.
Virtual calipers are available within the DANA 3D viewer for basic measurement functions (length, width, thickness, circumference, contour, etc.) of objects, and more advanced morphometric tools will be made available in the future. The points of measurement on the models can be saved images for future reference.
A data mining tool suite is in development. This will provide domain experts with a set of relationships not previously considered and from which new research directions could emerge. For instance, it would allow users to request a list of all relationships among a set of attributes with a predefined level of confidence.
2 Purpose and Value of DANA
2.1 Virtual Archaeology
3D computer visualizations are becoming more widely used in various scientific disciplines, vastly improving our means of understanding the world (e.g. Chen 1999; Durlach and Mavor 1994; Earnshaw and Vince 1995; Rosenblum et al. 1994). In the 1980s, a small number of archaeologists, both anthropological and classical, began to experiment with 3D graphics and virtual reality modeling for archaeology. The concept of "virtual archaeology" (VA) eventually emerged, marked most clearly by a seminal paper by Reilly (1991):
"The key concept is virtual, an allusion to a model, a replica, the notion that something can act as a surrogate or replacement for an original. In other words, it refers to a description of an archaeological formation or to simulated archaeological formation."
Uses of 3D visualization by archaeologists have predominantly aimed to create virtual reality environments that simulate archaeological sites or reconstructed buildings. Solid modeling techniques have typically been the means for virtual site reconstructions, some of which are highly conjectural while most are becoming increasingly evidential-based and "realistic" (better use of perspective, lighting, texture, see Chalmers and Stoddart 1996) constructions. Efforts in this area have gone from simple simulations to reconstructions, from simple time-step walk-through to complex fly-through and animated tours, with even more complex augmented reality tours on the horizon.
While projects in virtual archaeology are increasing, they are still comparatively uncommon and peripheral to modern archaeology. We believe this will change in the next few years as such projects are taken to new levels of complexity and realism. For a survey of the increase of such works over the last decade see Lockyear and Rahtz 1991, Higgins et al. 1996, Forte and Silioti 1997, Barcelo et al. 2000, Burenhult 2002. The attraction of VA is that it enables a more realistic and visually rich presentation of information. Furthermore, with 3D visualizations, archaeologists can now link artifacts to physical contexts of locality, environment and culture in a way that can enhance understanding of intra-site and inter-site relationships.
There have been fewer attempts at 3D modeling of artifacts or human remains than virtual site simulations. In the past, 3D artifacts have often been creations drawn in a CAD program and given surface texture treatment, rather than accurate representations of real artifacts. Through recent improvements, laser scanning and photogrammetry can now provide a means to produce metrically accurate 3D representations, or digital models, of nearly any artifact.
Even when accurate, 3D digital artifacts are produced, they are typically for novelty or special purpose (e.g. a specific display, or key artifacts from a specific site), but not for systematic collections management, and certainly not with the intent of making the models available as part of an archive. That is the logical next step, especially as the 3D modeling process is refined and the equipment becomes more affordable. This goal is precisely the purpose of DANA. The digital artifacts included in DANA can then be used for a wide variety for purposes, from individual study to embedding in virtual archaeology sites. Indeed, once operational, DANA will provide a valuable resource for cultural heritage management, research and education.
2.2 Cultural Heritage
Access to, and preservation of, the physical remains of human cultural and biological heritage are essential for an informed society. Unfortunately, access and preservation are subject to several potential problems (Clark et al. 2000; Clark et al. 2001):
- A growing "crisis in curation" (SAA 2000; NSF 2001; NPS 2000) resulting from increased recovery of artifacts and fossils over the last few decades and a critical shortage of storage space for those objects.
- Memory institutions, which are the repositories of human heritage materials, are relatively few in number and unevenly distributed around the globe, resulting in an overwhelming inequality of access.
- At memory institutions, exhibit space for public viewing of cultural heritage objects is always limited in comparison with the quantity of objects actually curated.
- Access to cultural heritage collections is typically severely limited, even for scholars, due to costs of travel to repositories; this creates even greater disparity of access between researchers based on financial resources.
- Many antiquities are fragile and should not be subject to repeated exposure and handling.
- Handling of culturally sensitive materials may not be appropriate or permissible.
In short, scholars, as well as the general public, too often lack the opportunity to explore the vast majority of objects that tell the story of what it means to be human.
Information technology provides the opportunity to help correct these problems. Laser scanners, or digitizers, and sophisticated software can be used to create 3D models of the external shape and surface characteristics of nearly any object. These models can be digitally stored and retrieved for viewing on a computer screen, or sent to others via a portable storage medium (CD ROM, DVD, Zip disk) or Internet connection. As digital representations, these materials can be used repeatedly for future studies and teaching without damage to the rare and sometimes delicate objects.
Moreover, model data can be archived, along with associated contextual information, in an openly available, Internet accessible, federation of distributed databases. The DANA interface provides users with the ability to access multiple databases simultaneously at a variety of locations, thereby significantly improving the ability of researchers to access richer and broader ranges of information about human heritage.
2.3 Research Potential
The digital models will be accompanied by comprehensive historical, referential, qualitative and quantitative data for each object in the database. The analyses made available through new and sophisticated software, data mining and accurate imaging will significantly improve studies of material remains and our understanding of history, prehistory, and human evolution. Digital 3D models permit:
- Precise linear and volumetric measurements of objects, some of which are difficult to make with traditional methods.
- Virtual dissections of an object for detailed examinations.
- Virtual model constructions so that distorted, missing, or ambiguous areas of artifacts or fossils can be simulated in multiple versions.
- Morphological comparisons of multiple specimens of the same type of object done either by superimposing multiple images in the same coordinate space, or by statistical analysis of coordinate data points.
- Consistency in measurement methods.
- Illustrations of methods, materials, measurement points and results of investigations.
Ultimately, DANA will provide user-friendly technology for Web-based collaboration between two or more remotely located individuals through access to virtual laboratory environments (ATL 2001a). Users from distant locations will be able to hold real-time online meetings and seminars in virtual environments for study and analysis of accurate 3D models.
2.4 Educational Benefits
DANA resources will enhance the ability of educators to draw upon a broad range of rich, anthropologically oriented digital content for use in the classroom. Educators will be able to select, assemble and use DANA content appropriate to their target audiences and teaching level, especially as that content is coupled with other information to create educational modules. Several examples follow:
- Elementary school teachers can select, assemble and use DANA content to develop and deliver topic-specific teaching modules, such as providing an overview of 1800s era Plains Village Indians life ways.
- In secondary or post-secondary settings the same content can be used to study basic village social organization.
- For post-secondary education, the content may be used in modules addressing kinship, ritual, and other more advanced topics.
- The same content that interests a social studies teacher or anthropology professor might also be of interest to someone teaching history, Native American studies, folk arts, or other related subjects.
- Future developments include a virtual teaching lab where students will be introduced to various types of material analyses and data recording on digital 3D models drawn from DANA.
- As future virtual reality models of archaeological sites are created, they can be populated with digital artifacts drawn from DANA, thus further contextualizing the artifacts. One such project under development by the ATL is the Virtual Archaeologist, which is an immersive virtual environment (IVE) for education (Slator et al. 2001). Virtual Archaeologist is being developed along the lines of similar IVEs that have been created by the NDSU World Wide Web Instructional Committee (WWWIC 1998-2000).
- Museums will be able to create virtual exhibits in kiosks or on large screens that can, at the touch of a finger, display a larger portion of their collection than is otherwise exhibitable (Kerendine 1997; Kadobayashi et al. 2000; Purcell 1996;Russo 1998;Terras 1999).
3 DANA System
The system architecture employed at NDSU involves multiple servers, but that level of internal distributing is not required for participating institutions. Interoperability of network participants and connection with the much larger NSDL network is made possible through the use of open standards, metadata and XML.
The prototype system developed for DANA uses a set of three servlet types to handle archive queries. The system is described in this paper. However, we are currently in our second phase of system development during which we are exploring options for redesigning aspects of this system. As currently designed, a Manager Servlet controls a large region and answers broad queries. An Entry Servlet provides a point of entry into the system and stores persistent user information. A Data Servlet contains information about each object and will be the most common type of servlet.
A user can access DANA from any designated entry point by way of the servlet system, which provides access to all participating databases throughout the distributed network. The DANA data entry tool will allow participants to correlate their database or spreadsheet content with DANA templates. Within DANA, a wide range of searches can be carried out. Data can be retrieved, and 3D or 2D images can be viewed. A set of data mining tools is currently under development at the ATL to facilitate data extraction and interpretation.
3.1 DANA Architecture at NDSU
At present, the DANA system at NDSU incorporates the campus network and backbone, the NDSU Oracle 8i database, the ATL Research Storage (RESSTOR) server for image and model files storage and retrieval, and the NDSU Web server for application operation. The ATL also uses test systems consisting of the Apache Web Server, Apache Tomcat servlet container, PostgreSQL, and a Control Version System (CVS). It is important to note, however, that minimum requirements for participating institutions are simply a servlet container and a compliant database.
3.2 Interoperability and Information Technology
To establish network interoperability between the participating institutions' RDBMSs and file servers, we are using open standards such as SQL and XML (Extensible Markup Language) combined with new Java technologies. XML, a platform-independent and license-free text format, provides a mark-up language to contain data in a way that facilitates data sharing between computer processes. As such, it will provide a means of standardization for Internet distribution of data.
The NSDL program has adopted an extended Dublin Core Metadata Initiative (DCMI) metadata model for collections level descriptions (Hillman 2001).
Elements, qualifiers and extensions of Dublin Core (DC), Dublin Core for Education (DC-Ed), and several elements from the Content Standard for Digital Geospatial Metadata (CSDGM) have been added to the NSDL metadata standard for collection level descriptions. DC interfaces for the Resource Description Framework (RDF) have been developed as a World Wide Web Consortium (W3C) recommendation, and will also be used in the DANA project. Use of these metadata schemes will facilitate Internet access, content query and information retrieval for users of the NSDL Core Integration System (CIS) Portal at Cornell University.
3.3 DANA Servlets
The DANA system is currently designed to use three levels of Java servlets to handle database queries with the remote servers: Manager, Entry and Data (Figure 1). This allows us to load the JDBC drivers at the servlet level for the appropriate database systems under that servlet's control. The connection to the RDBMS is implemented through a connection pool that can be set to open connections as needed. As a result, the pool will not use more connections than are required at any given time, maximizing efficiency and cost-effectiveness by not tying up more licenses than needed.
3.3.1 Manager Servlets
A Manager Servlet acts as a control mechanism for the system. Only one Manager is required for a large geographical region, so a small number of managers will be needed for the entire network. By using a small number of distributed Managers we can provide redundancy while also improving response time. A Manager Servlet contains user information that is mirrored across the system.
The Manager servlets answer broad queries, such as what materials are in the system, specimen types for a specific list of materials, and the location of collections. All other servlets register with the Manager servlets before the system can be accessed for purposes of system query, data retrieval, or data entry. Only after those three items have been verified will the requesting servlet be brought online. This will prevent unauthorized systems from participating in or accessing the system and further prevent, for example, insertion of false or invalid information. The managers also detect active servlets and will notify the appropriate system administrator when a servlet is not performing as expected.
3.3.2 Entry Servlets
The Entry Servlet acts as a point of entry into the system. All clients will connect through an Entry Servlet. The same user can use different entry points, even if s/he did not register initially through that entry point. The Entry Servlet will contain persistent information related to users. If an Entry Servlet goes down, the user will still be able to enter the system, although s/he will not have access to their persistent, personal data.
3.3.3 Data Servlets
The Data Servlet provides direct access to one or more RDBMSs containing object data. The Entry Servlet determines what information any particular Data Servlet will provide and sends the client as well as the Data Servlet a "key" that references those data. The Client uses the key to retrieve the data from the Data servlets. This minimizes network traffic and potential load on the Entry servlets.
3.3.4 DANA Servlets Summary
We will be experimenting with various adaptations of this basic system to find the best mix of efficiency and flexibility. Some of those adaptations will use XML and AnthML. One of the fields available to the client is the URL for the image or images of each artifact identified. The client will retrieve the image directly from the server on which it is stored. The DANA system is designed to scale well, from running entirely on one computer to containing any number of sites, each of which could employ a multi-server architecture with up to four servers (even more in a high availability cluster scheme), each with different functionality.
3.4 Participating Databases
The primary task of the institutions providing content for DANA will be archiving of data in textual and visual forms. We have developed a data definition format for test-case objects, and we will continue our development of a prototype search interface and retrieval mechanism that make these data accessible to anyone with an Internet connection and Web browser. While the inclusion of 3D models of artifacts is a significant and unique part of DANA, contributions of textual data, photographs, maps, etc., will also be sought from participants in the network.
The database tables will be available for certified users to download and use to create new artifact tables with relative ease. The local system administrator may grant "write access" to qualified individuals for his/her database. Tables can be created and data entered through either traditional methods or the DANA Data Entry tool.
3.5 DANA Data Entry Tool
We are developing a Data Entry tool for the system, which provides options for either entering data to an existing table or letting the user create a new table. The data entry tool will extract data from an existing database, a comma separated variable file (*.CSV format), or any other delimited (e.g. '|' or ':' or '/' , etc.) text file, and will aid the user in creating and populating tables. A certified content provider logs into DANA, and s/he can choose any data file (e.g. .CSV file) from the file dialog box and view it as a table. The user can then browse through and view the existing tables, which are available as templates. If none of the existing tables match the data, the user can create a new table either by selecting the fields from existing tables or by inserting a new column. The user can then align the columns of his/her data with those of the template by dragging and dropping. Once the user is satisfied with the correspondence of columns, the data can be inserted into the existing or newly created table. The data entry interface facilitates platform-independent, database table creation, and it can be used to transform and migrate proprietary RDBMS formats to any compatible RDBMS format. We are also exploring the use of XML as an alternative method.
3.6 Data Mining
Data mining (knowledge discovery in databases) has traditionally been defined as the extraction of non-trivial, implicit, previously unknown, and potentially useful information or patterns from data in large databases (Han and Kamber 2000). Data mining will be an important component for research done through DANA. Data mining will not only be able to find new relations between and within data sets, it will be able to provide insight as to whether existing classification systems are complete or accurate.
3.6.1 DANA as a Source of Data
The DANA federation of databases will provide a rich warehouse of data that can be mined for the kinds of knowledge that domain expert users may desire. Automated data mining and knowledge-discovery support tools can be developed specifically for this purpose. In addition, these tools can be used to examine data in new and interesting ways. We are working to create a data mining tool suite for DANA that will allow users to request, for instance, a list of all relationships among a set of attributes with a predefined level of confidence. Such data mining features could provide domain experts with a set of relationships not previously considered and from which new research directions could emerge.
3.6.2 Potential Data Mining Applications
There are a wide variety of potential applications of data mining in the DANA architecture. Implementation will be at several levels. DANA contains two categories of data for data mining: descriptive attributes and visual files. Descriptive data about artifacts in DANA can be compared and clustered to identify types and categories or rules about relationships between data. While the initial data mining applications will focus on the descriptive data, it is hoped that at a later stage the 3D models themselves can be mined directly.
3.6.3 Previous and Prototype Data Mining Efforts
Previous data mining efforts at NDSU include the SMILEY (Signature Miner and Interface Language for Earth Yield; SMILEY 2001) suite. This work will be used as an initial approach for DANA data mining (Figure 2).
A Web-based remote sensing imager, data-mining system, SMILEY provides a general interface to access, display and manipulate massive amounts of spatial data through any common Web browser. It was designed as a system for accessing, distributing and analyzing spatial data in order to address identified needs in education, precision agriculture, environmental monitoring, genetic research, and more. A preliminary prototype for data mining of metric and non-metric Samoan adze attributes using parallel coordinates (Inselberg and Dimsdale 1994) is available for viewing at http://midas-10.cs.ndsu.nodak.edu/data/pc_demo/. The prototype module using Straight Parallel Coordinates with Samoan Adze Data has been developed using Java and will be integrated into DANA (Figure 3).
A prototype approach projecting similar data on to a 2D polygon to identify class separations has been created using VRML and can be viewed using DANA's 3D viewer. This demo of the jewel diagram (Juell and Jockheck, in press) can be accessed at http://jockheck.northern.edu/vrml2/demo12a.html but requires a VRML viewer (see Computer Associates International) (Figure 4).
Figure 4. Jewel diagram showing separation of type 1 and type 3 Samoan adzes based on attribute data(full size)
3.7 Conceptual and Practical Visualization
When working with anthropological data, researchers and students typically find themselves working with extremely complex multi-dimensional data. This is often difficult to organize into a reasonably simple and understandable model. For this reason, we are developing graphic modeling tools to assist users in creating practical visualizations of both the facts and abstract concepts represented in the data. Visual presentations will enhance the user's ability to understand the underlying information.
4 Using DANA
To access DANA, users must download and install a Java-based application. Once the application software is installed, users are able to search DANA at differing levels of specificity. Queries will generate a list of objects that match the search variables, and list items can then be selected for specific data retrieval. When 3D models are brought into the DANA viewer, virtual calipers are available for basic measurements. Those measurements, along with image snapshots and other observations, can be stored in a personal notebook.
4.1 Required Downloads
To use DANA one must download Sun Microsystems Java Web Start (JWS) (1.0.1) software in order to install the DANA client application (currently DANA 0.2). A tutorial (ATL 2001b) is available. During the installation, JWS will inspect the user's computer to verify that the Java 1.3 runtime environment (JRE) is installed; if not, JWS performs the JRE installation automatically. Once the JWS installation is complete, the user returns to the ATL Web page where s/he is provided with the location of the Java3D JRE download, which is required to implement the 3D viewer in the DANA client application and is available for Linux, Solaris and Windows. Although Java 3D is not supported on Macintosh operating systems, the other DANA applications in the Java Web Start client work on Mac OS X (Mac OS 10).
4.2 Searching DANA
A valid login presents the user with the search interface. For a detailed set of instructions on interface functionality and operation, the user may consult the online tutorial or the DANA client's help files.
The interface will eventually allow searches by geography, cultural affiliation, site code (number and/or name), taxonomy (for biological materials), and material, or some combination of these, but in the current version, initial browsing is geographical. Drop-down windows enable a geographical search through six levels of increasing spatial specificity: global regions; global subregions; individual country; and primary, secondary, and tertiary levels appropriate to the country (e.g. in the US, state, county and township). When the desired geographic level is reached, the user can continue the search by material category (e.g. stone, ceramic, etc.) and artifact type (e.g. adze) (Figure 5).
The search is then refined by querying the database for some set of attributes (e.g. a specific cross-section, a specified range of dimensions, etc.). The search returns the result as one list of object IDs for DANA entries that meet the search specifications. From the query-result list, the user will be able to select the specific artifact(s) for which s/he wishes detailed information by accessing specific database tables: all metric or non-metric attributes; provenance; chronology; documentation (e.g. location of artifact, relevant reports or publications, etc.); and images.
4.3 Data Retrieval
The visualization interface will be able to retrieve 3D models (e.g. VRML or WRL format models), with or without actual color (depending on original scan and the size of the data file), or 2D images of the artifacts (Figure 6).
Ultimately, we will accommodate the users' access mode (Internet 2, Ethernet, modem), computer capabilities (RAM, speed, graphics card), and purpose (e.g. detailed analysis, use in a VR model, or general observation). For full models the file sizes are quite large (e.g. 20-30 MB, or larger), but the resolution is so precise as to allow measurements at the sub-millimeter level. We plan to employ some combination of compression software and file decimation to deliver the large files.
4.4 Measurement Functions
The ability to make accurate measurements from the 3D images is a significant aspect of our project. We have developed the first round of "virtual calipers" that permit measurements of linear distances, either along the surface (contour) or through the body of the artifact (cord), and circumference. More sophisticated shape analysis and measurement tools will eventually be made available as options. We encourage collaboration in the development of those tools, and in the refinement of our viewer.
Users will be able to store their measurements and record other observations in a password-protected personal "notebook" that will persist between sessions. At any point, images on the screen can be saved to the notebook or some other file, emailed or printed. This allows the user to keep detailed documentation of the objects examined, tools used, the measurement points, and other important information.
5 Collaboration and Participation
DANA implementation will involve efforts in multidisciplinary research (e.g. archaeologists, physical anthropologists, ethnologists, museologists, archivists, computer scientists and mathematicians), multi-institutional collaboration, and multi-national cooperation. Consequently, the project team has been actively seeking participants at local, regional, national and international domain levels.
5.1 Collaboration
There are differing levels of collaboration in the DANA network development. It is expected that staff from institutions that are major participants--with a Manager Servlet--will be in active collaboration with comparable units in the development and expansion of DANA. At the same time, however, other collaborations may not entail network participation. For example, we hope to collaborate with other programmers to develop measurement tools or refinements of the 3D viewer used in DANA, or development and implementation of AnthML, even though those collaborators may not be otherwise involved. We may also collaborate with developers of similar digital networks that focus on other scholarly domains. People interested in participating and collaborating in DANA are encouraged to contact the ATL.
5.2 Participation
Participation in DANA can take place at many levels. Content providers may opt to be major participants, along with NDSU, which is to say that they maintain a Manager Servlet as well as Entry and Data servlets. Very few such nodes will be required, perhaps one for each large region of the US, for example. Other institutions may provide a point of entry to the network by maintaining Entry and Data servlets, while other participants may employ only the Data Servlet, thereby sharing their content with all users.
6 Quality Control and Sustainability
For DANA to work, a large number of research, learning and memory institutions must work cooperatively to create an integrated system. There is a clear need for standards to ensure that fundamental database information is collected in a consistent format by all archive contributors, thereby promoting interoperability between databases. To that end, DANA developers will work with collaborators around the globe to develop a standardized vocabulary of cultural heritage objects--along with a glossary and thesaurus for terminology--to facilitate cataloging and electronic search and retrieval.
A DANA Advisory Board (DAB) will be established consisting of informed individuals who are participating in DANA at some level. The DAB will certify institutions that wish to participate in DANA, and those institutions can then, in turn, certify individuals to enter content. The DAB will also oversee appropriateness of content for DANA, and will seek to bring new content providers into the network.
The model that we propose for DANA is one of vested ownership rather than hierarchical control. We anticipate that different participating intuitions will take the lead in materials for which they have high expertise and suitable collections. Again, DANA is a network of databases, not a single data repository. The only restriction on content should be quality of the data.
As a distributed network, the costs of maintaining the DANA system are also distributed, so that no one institution must bear the brunt of the costs for sustaining the network. The costs of developing a 3D scanning and modeling system vary depending on the quality of the equipment purchased. We anticipate that in the future costs will drop as hardware and software manufacturers develop new products, and due to the development of more cost-effective ways for system operation.
Not every participating institution will need, or wish, to develop its own digitizing laboratory. Institutions and organizations with materials they would like archived can have that done for a relatively small fee by labs such as the ATL at NDSU. Also, people with few resources (e.g. slow database server, low bandwidth) are able to participate in DANA because the high-resource servers can cache the data from smaller servers. This will effectively decrease the traffic on the slower machines and enhance the response time of the entire system. Those sites with large databases could decide to participate as caching or non-caching servers.
7 Conclusion
DANA development integrates a number of innovative technologies into a novel application. The benefits promised by this project are both immediate and far-reaching in terms of serving scholars in archaeology and anthropology, in providing access to vast networked datasets, and in supporting scientific inquiry in a way that largely eliminates the barriers of time and distance.
A distributed network such as DANA is clearly in keeping with the current trend for museums, libraries and universities to use the Internet and IT as an alternative avenue for making scientific collections more accessible and useful to a much wider audience than the brick and mortar facilities can service (Science 2000). Indeed, we maintain that DANA, or something very much like it, represents the future of cultural heritage management for research and education. Furthermore, an Internet-accessible archive will improve data access and reduce inequalities in research and teaching resources.