3.8 Integration of OA with Repositories
This section is divided into two sub-sections. First one is intended to describe, integration with existing search services. Second sub section will be concentrated on integration with research administrative systems.
3.8.1 Integration with Existing Search Services
An institution can facilitate different kinds of search services- e-journals search, external database search, library OPAC search, IR search etc. which are most comprehensive and common facilities provided by institutions. All these services in a typical library setup are available through different user interfaces. It is less than an ideal situation as far retrieval efficiencies are concerned. To avoid this situation, institution may facilitate a single window search interface by using discovery tools or by using custom search engines. The way by which an institute can provide single window search interface may be referred to as integration. Almost each and every institution provides more than two services to their end-users. At the same time, they may provide information relating to end-users’ search query from their internal database (includes Library OPAC, Institutional repository, e-journal database) and from external databases (e.g. open access database like BASE or subscribed database like Scopus). It becomes much awkward and monotonous to jump from one search interface to another to retrieve any information regarding any search query. In order to consolidate this search, a mechanism must be there to search the entire external and internal databases through a single window search interface. It is quite simple to do this by using tools like Custom Search Engine, Discovery tools etc. Before creed to mechanisms we should be little bit acquainted with these two terms.
Custom Search Engine (CSE)
Custom search services are facilities to use established search engines to organize and index selected Web entities though API (Application Program Interface). This may be applied to index OA journals and OA repositories. Each such custom search service has URL and search only resources included in it. Finally, the custom search services may be integrated with local repository or local user interface to provide a single-window facility to search local OA repository as well as global OA repositories. The major search services like Google, Yahoo etc. provide custom search mechanisms to empower the users for search. These facilities allow searchers to index a set of selected websites. Each custom search interface provides URL (Uniform Resource Locator) to ensure global access. The process of integration is straight forward and includes steps mentioned below:
- Development of CSE on a given area: You may try developing a CSE by utilizing Google API for open access LIS journals.
- Widget generation: Widget is an application, or a component of an interface, that enables a user to perform a function or access a service. Google CSE allows users to generate Widget in the form HTML and Java code. You may generate Widget for your CSE by simply clicking the appropriate button.
- Widget integration: The next step is to integrate Widget for your CSE in existing library systems (like Web-OAPC) or OAR search interface.
- Testing and Debugging: Finally, after testing and debugging (if necessary), the product is ready for end users
Discovery tools, powered by federated search mechanisms, allow users to perform concurrent searching in the library catalogue (metadata level), journal articles (full-text level), electronic theses and dissertations, consortia databases, public web, open access repositories, union catalogues etc. through a single-search interface with a set of feature-rich tools to support users. In simple words, a web-scale discovery services allow users to search local and remote databases, open access and commercial knowledge from a pre-harvested single central index. The unified interface is a big boost for users as they no longer need to choose a specific search tool to begin their search. These tools are available commercially (e.g. EBSCO Discovery Service) and also as open source products (such as VuFind, SOPAC, Blacklight, OpenBib etc).
3.8.2 Integration with Research Administrative Systems
Researchers register their work by giving metadata like title of a research work, name of researcher etc. under an institutional system. It may be under university system or under any funder agency. So institutions should maintain all records for official as well as academic purposes. In the Research Administrative System, at the time of submission of research thesis, it will automatically be uploaded in institutional repository also. In this context we may have a short look on CRIS/OAR Interoperability Project. Current Research Information and Open Access Repositories (CRIS/OAR) transfers metadata of publications automatically from research information system to institutional repository with option (from authors) to integrate full-text resources. It aims to achieve grand unification of research administration needs and OA repositories.
CRIS/OAR Interoperability Project
This project started in January 2009 and outcomes were presented in October 2010. The latest release of open source repository management software Dspace supports CRIS/OAR. This allows Dspace repository population by automatically transferring metadata-only records from the CRIS to the OAR and asking authors to attach the appropriate full-text versions of the works to the records in the repository.
1. Use Google Custom Search (https://www.google.com/cse/) and develop a search engine to search open access contents across a specific discipline. Register open access journals (see http://www.doaj.org/) and open access repositories (see http://www.opendoar.org/index.html and http://roar.eprints.org/) as targets.
2. Use any open source harvesting software (e.g. PKP Harvester) to create a federated search interface for any five OAI-PMH compliant IRs in an area (discipline) of your choice.