The short description: jake (Jointly Administered Knowledge Environment) helps you find an electronic copy of a dcoument that you know that you want to find. This hardly qualifies it as a "knowledge managment" tool.
jake is designed to work with various eletronic publishers, to help users find documents that they may then need to pay to acces. The metadata managed by jake is provided by the various publishers. jake is not intended to replicate or redistribute any actual e-resource content; rather it is to facilitate access to and integration between contents of many e-resources by appropriately relating metadata about e-resources and their content.
Jake is an open source project which can be found on sourceforge (which could be considered a knowledge management tool itself). Jake is in a state of development and a finished commercial quality product may never be completed. Jake is written in PHP and Perl, and is covered under the GNU Public License (GPL).
A working demo of jake can be found at http://jake.med.yale.edu/
| Back to Top |
jake is a reference source which makes finding, managing, and linking online journals and journal articles easier for students, researchers, and librarians. jake does this by managing online resource metadata with a database union list, title authority control, and linking tools, as well as making it easy to add a local holdings layer.
The user of jake enters a partial name of a journal, and specifics about a desired article, and jake returns a list of locations where the document can be obtained, and it what formats.
More Info: About jake
| Back to Top |
jake creates metadata from information provided by the electronic publishers (or manually provided for local collections). Scripts are provided for converting summary info in common formats to jakes internal metadata. These scripts are easily modified to accomodate other formats. There is a standard minimum set of data for source lists to be added to jake. This can be augmented by additional feilds for local collections.
jake uses SQL for the metadata back end.
| Back to Top |
jakes metadata is stored as data in mySQL.
searches for articles or journal names can use partial words and wildcards.
jake allows searching by keywords, both secifically listed and in summary abstracts, and by citations or other references to the document.
No sophisticated knowledge managment tools are provided to group together titles on similar subjects. There are no provisions for the use of an ontology or for expanded searches using a thesarus.
| Back to Top |
The found e-resources for the searched for document are presented in tabular form allowing the user to verify the document and choose the most appropriate publisher and format.
While the table shows some relationships between documents and presents icons for different document formats, there is no grouping of the data to help organize the presentation.
| Back to Top |
The only addition to simple keyword search that jake provides is keyword expansion, where only the first part of each word needs to be entered. Common abbreviations for journal names are also expanded.
| Back to Top |
jake is set up so that the meta data can be collaboratively expanded, but no provisions are made for finding whether a particular article is relevant or if an entry is accurate.
| Back to Top |
jake is not only the software for accessing the metadata, it is also the meta data itself. The metadata is expected to grow as different libraries and publishers contribute and share their data.
jake provides tools for merging metadata from different sources. However, it does so in a trusting fashion, with limited checking to verify the collaboratively contributed data.
| Back to Top |
jake was created to capture the collective knowledge of how to access specific document electronically. jake is not meant to be proprietary, but a collaborative effort over many libraries.
| Back to Top |
jake may be obviated by other efforts, such as CiteSeer which are also open source projects.
more and more scientific literature is becoming available freely on the web, both from publishers directly and from the the authors themselves, sometimes in draft form. This makes the additional redirection of jake to a publishers pay site uneccessary.