ASF ASF logo HOME | MIRROR | BROWSE | SEARCH | E-MAIL |
Updated 15 SEP 1998 Security and Privacy Notice | URL:/about/faq.html

FAQ (Frequently Asked Questions)

Introduction to the Advanced Search Facility (ASF)


What is the ASF?

The ASF is about software tools that perform a number of resource location tasks.

The goal of ASF is complementary to the goal of the Global Information Locator Service (GILS)--to make it easy for people to find information of all kinds, in all media, in all languages, and over time.

Specifically, ASF implements a scaleable, distributed search and retrieval capability and permits the development of intelligent agents to query large collections of mixed-mode data (e.g., text, geospatial metadata, and bibliographic references). GILS provides two critical components--a full-featured, powerful search and retrieval protocol; and a robust, record-level metadata object that can clearly and concisely describe a document, data set, image, or other network information entity.

The ASF software implements an architecture which isolates the various components, so that their interfaces are well-defined, and so that they can be replaced by third-party components written to use the same interface.


What do you mean by saying ASF is an architecture?

The ASF architecture defines how the various components communicate with each other. By building the ASF software in a modular fashion, it becomes possible for other software developers to extend or enhance the functionality of ASF to suit their needs. As long as the requirements for GILS-compliance are met, the underlying implementations are free to change.

For example, you might want to replace the default, free text search engine in ASF with one of your own choosing. There is a search interface which can be configured to support other search engines. The appearance of the searches to the outside user will not change - only the functionality of how the search engine answers queries will be different.

Who verifies that software is "ASF-compliant"?

Compliance with ASF is defined at the standard interfaces in the ASF architecture. A test suite for these interfaces is being defined.

What does ASF offer to me as an Information Community organizer or participant?

ASF offers the opportunity to support a widespread, commonly accepted search interface to your collection of online resources. It offers the ability to let users search your site in a standardized way, and it offers the ability for you to refer users to other sources of relevant information, just as they refer users to you. This is a way to help organizations make their information easy to find.


Is ASF just for networked information?

Information providers can describe anything with ASF--not only books and datasets but people, events, meetings, artifacts, rocks.... For networked information, ASF supports "hyperlinks" for network access to the resource described or related resources.

How much does it cost to put up a ASF-compliant server?

Freeware implementations of ASF are available for some Unix platforms. Of course, these may be more difficult to administer, may have minimal documentation, and typically do not have formal support.
Commercial ASF-compliant software starts at about $6,000 for a small Alta Vista implementation. Additional options are available for sophisticated database support such as Informix, Intranet knowledge management such as Fulcrum, or library support with products such as SIRSI and OCLC FirstSearch. (For company contacts, see </partners/commercial.html> )

Where can I get ASF software?

Pointers to sources for ASF software are maintained at </technical/index.html>.

Where can I find more information about ASF?

More information about ASF is at </>.

Is there an ASF mailing list?

The Coalition for Networked Information (CNI) is hosting a mailing list for ASF. As described at </about/forum.html> you can subscribe to the list by sending e-mail to listproc@cni.org with a single line in the message body:
SUBSCRIBE ASF <your name>

Where can I find more information about GILS?

More information about the Global Information Locator Service (GILS) is at <http://www.gils.net/>.

ASF Freeware


What is ASF Freeware?

ASF is a freely available software distribution that combines several open standards to implement a GILS-compliant resource location system for information communities. More precisely, ASF is the architectural model for how various components of the system fit together. The individual components themselves may be interchanged with other ASF-compatible software or upgraded to commercial implementations.

ASF provides an integrated system for searching documents and document metadata via HTTP and ISO 23950. It also provides a distributed resource discovery mechanism that can integrate searching among a community of ASF servers.

ASF freeware supports automatic indexing of the content of World Wide Web sites. An ASF Gatherer Service extracts and compiles documents using GILS-compliant search interfaces as well as hyperlinks and directories. A GILS-compliant ASF query service supports full-text, fielded Boolean, and spatial queries using a combination of human-crafted and automatically generated metadata.

Who is involved in developing ASF Freeware?

The following organizations are involved in the ASF project


Where can I find the ASF Freeware?

See the ASF freeware distribution page.

What high-level components are included in the ASF freeware?

The ASF node software includes:


What platforms has the ASF software been tested on?

We have confirmed that asf-0.3a and higher build and run under:


What will my platform need pre-installed before building and using the ASF node software?

Related Technologies


How does ASF relate to the Common Indexing Protocol (CIP)?

The search referral component of ASF is based on the Common Indexing Protocol (CIP). CIP is an Internet RFC produced by the IETF (Internet Engineering Task Force) FIND working group <http://www.ietf.org/html.charters/find-charter.html>. The FIND working group focused on Internet directory services, including X.500 and Whois++. CIP is intended for use by directory services when passing indexing information so that a user query to one server could also result in referral to another server based on the exchanged index information. An example of indexing information is a word occurrence metric such as a "centroid" (for example, see <http://www.ilrt.bris.ac.uk/roads/cross/> ).

How does ASF relate to the Dublin Core?

Dublin Core is a set of definitions (semantics) for some common metadata elements. The further development of Dublin Core is now tied to the work on Resource Description Framework underway in the World Wide Web Consortium (W3C).

ASF and many other operational implementations demonstrate that GILS-compliant search is used very successfully with Dublin Core. The fifteen unqualified Dublin Core elements are mapped to GILS by the Library of Congress Network Development and MARC (Machine Readable Cataloging) Standards Office (see <http://www.loc.gov/marc/dccross.html>). Dublin Core does not specify syntax (although there is a W3C proposed convention for how to represent Dublin Core elements within HTML).


How does ASF relate to the Extensible Markup Language (XML)?

Extensible Markup Language (XML) was developed by the World Wide Web Consortium as a version of Standard Generalized Markup Language (SGML) for electronic publishing and data interchange. (see <http://www.w3.org/XML/> )

Designed for Web content providers, XML may become a common metadata representation in Web objects and elsewhere. As a standard for metadata representation, XML is complementary to ASF in the same way as SGML. In the World Wide Web Consortium work on Resource Description Framework, a further possibility is emerging in using XML as a way to specify multiple metadata schemes and place the various sets of semantics into a commonly accessible and automated registry. Standards work in this direction is clearly part of the evolution of ASF.


How does ASF relate to Harvest?

Harvest focuses primarily on the process of gathering Web resources, an important component in typical ASF implementations. ASF does not mandate any particular software among the many varieties of harvesters, crawlers, robots, gatherers, and so on. Instead, ASF defines the architectural interface needed to support such techniques for searching.

How does ASF relate to the Lightweight Directory Access Protocol (LDAP) and X.500?

The X.500 family of standards provide a robust series of mechanisms for management of and access to directories such as electronic mail and other addresses. The X.500 Directory Access Protocol, and its "lightweight" version for the Internet, incorporates a query service for the directory. Because the LDAP query interface can be regarded as a special case of search, one can provide a gateway for ASF using the Z39.50/GILS interface to access data from an X.500 distributed hierarchical database.

How does ASF relate to the MARC standard?

The MARC (Machine Readable Cataloging) standard provides a combination of syntax and semantics for a range of applications that are primarily bibliographic. (Semantics are more properly described separately in cataloging rules such as the Anglo-American Cataloging Rules.)

ASF incorporates GILS which adopted MARC semantics for the elements used in locator records. In fact, a one-to-one correspondence of GILS elements to MARC tagged elements is maintained in the GILS Profile. (see <http://www.gils.net/prof_v2.html#annex_b> )

How does ASF relate to the W3C Resource Description Framework (RDF)?

RDF <http://www.w3.org/Metadata/RDF/> is being designed as an infrastructure to support metadata across many web-based activities, with different application communities each defining the metadata property set that best serves their needs. RDF intends to provide a uniform and interoperable means for exchanging metadata between programs and across the Web.

So far, RDF has defined XML (Extensible Markup Language) as the transfer syntax and has drafted a schema definition. The overlap with ASF will be primarily in the RDF schema(s) and any subsequent work on services, especially search.

How does ASF relate to the Standard Generalized Markup Language (SGML)?

ASF uses XML, a variant of SGML, as the preferred syntax for ASF locator records.

How does ASF relate to the Stanford Protocol proposal?

Like ASF, the Stanford Protocol Proposal proposed GILS elements for metadata (see <http://www-db.stanford.edu/~gravano/starts_home.html> ) and the proposed search referral design also was quite similar to ASF.

How does ASF relate to SQL databases?

SQL (Structured Query Language) operates on relational databases. A relational database is one of the mechanisms that can be used to store locator records behind a GILS-compliant server as part of an ASF implementation.

How does ASF relate to WAIS?

Wide Area Information Servers (WAIS) was created using what is now the obsolete version 1 of Z39.50, documented as Z39.50-1988. ASF incorporates the GILS Profile, which requires Z39.50 version 2 or version 3. The commercial version of WAIS was updated to be GILS-compliant, but the freeware version of WAIS has not been updated.

Users of WAIS are strongly encouraged to upgrade to GILS and ASF. In addition to the ASF freeware, a commercial software upgrade path is available from Blue Angel Technologies

How does ASF relate to Whois++?

The search referral component of ASF is based on the Whois++ protocol but the ASF freeware servers cannot respond to Whois++ clients. The ASF freeware does not itself provide a Whois++ client.