Using Content on the Asia Think Tanks OAR

The site offers two primary means of finding content:

  • BROWSE—The browse function is available in the upper left-hand side of the site. Users can browse through the content by collection, author, date published, themes or titles. The browse function is useful when a user simply wishes to find out what content is available.
  • SEARCH—The search function is available on the upper right-hand corner of the site. Users of the OAR can search in a similar way to web search engines and can search content by collection, title, or keyword. A list of publications that match the search terms will be displayed. Users can further refine the search results using the filters within the advanced search or by browsing using the faceted terms on the right-hand side such as author, theme, subject, topic, date, volume, series, abstract, or location.


The ATT OAR provides additional functionality such as:

  • a permanent uniform resource identifier, which is a permanent address for digital objects on the internet and will allow users to cite the reference back to the OAR site;
  • links to citations on Google Scholar;
  • a special icon highlighting items that are marked by ATT as Major Knowledge Products;
  • a list of related publications when users are viewing item pages; and
  • links to data sets used by the Asian Development Bank and other partners.


The ATT OAR supports the Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH). This allows for download of the metadata either for the entire repository, a specific collection or an individual item.

Other OAI-PMH-compliant repositories can harvest metadata in the OAR. Once the metadata from the OAR has been merged into another repository, users on those sites will be able to search, discover, and click through to access all publications.


Any OAI-PMH-compliant repository can harvest specific publications or a subset of the publications available on the OAR. This is done through “virtual sets” based on predefined search queries and/or filters.

For support in downloading subsets of the metadata, please contact


Retrieving content programmatically from the OAR can be achieved via its REST API. A query builder tool is available on the following URL:

The API offers several endpoints enabling users to retrieve different types of information:

  • Communities (information about the repository’s architecture)
  • Collections (containers for publications)
  • Item (individual publication)
  • Bitstream (individual files)

Each endpoint enables to retrieve a list of all instances of each “object”, or a single one (based on its internal identifier).

An additional endpoint “handle” allows users to get individual objects based on their Handle public identifier (prefix + suffix).

Finally, an extra endpoint was added, which can be used to search the repository through the REST API as one would search DSpace using discovery (advanced search module of DSpace). This endpoint takes the following parameters:

fields (mandatory): 
This is a combination of the metadatafield and the values that is to be searched. These fields have to conform to the format: "metadatafield:value" 1. Multiple metadatafields (and their values) can be used in the search, these have to be separated by ";"
This parameter allows the user to specify which extra metadata has to be given for the returned items. If nothing is presented, the default "metadata" is used. Depending on the parameter, information about the collection, community, bitstreams etc can be expanded
The maximum number of results that should be returned (default = 10)
The number of results that should be SKIPPED in the returning list of results (The "limit" parameter will then take only the items after this start into account
Specifying which fields to use for sorting. Possible values: dc.title, and
Order allows the user to specify how to sort the returning results. Possible values: asc, desc (default = asc)
This is the handle of the collection or community that should be used to search in. If this parameter is omitted, the search is done over the entire repository

1 Exact search: All metadata fields can also be queried by one of their exact values. When - for example - requesting all items with dc.type "Papers", a normal query for "Papers" would also return items with dc.type "Working Papers". When instead requesting dc.type_exact:"Papers"; only items with the exact string "Papers" in their metadata field dc.type will be shown.


Note: results are always displayed 10 by 10 by default, make sure to use the “offset” and “limit” parameters to browse through them.