Text Searching

The Case Study database is linked to a powerful indexed text search engine, which allows you to search for relevant case studies using a sophisticated search query syntax (see Lucene search help for more details). The basic principle on which the search engine operates is to return a set of results ranked according to how closely they meet your search criteria.

Simple Searches

The simplest form of search is to enter a single-word search term, e.g. diabetes.  The search engine will return a set of results listing all case studies in which this term occurs in any of the document sections. The results will be ranked according to the number of occurrences of the search term in the document.

Typically a single-word search term will result in many results, which may not necessarily be that helpful. The list of results is accompanied by a set of filters (to the right of the screen) that allow you to explore and focus the case study search results by institution, discipline, location and so on. Entering multiple search terms, e.g. diabetes research, delivers a more targeted search. The search engine will now return all case studies containing either search term, but with a higher ranking being awarded to case studies that contain both words.

A phrase, e.g. "diabetes patients" with the search terms enclosed in “ “ quotation marks specifies that the words are to be treated as a single phrase. This gives even more precise targeting.

Boolean Searches

The search query implies by default the presence of an ‘OR’ between multiple search terms. This will therefore return all records matching either term. If only cases studies matching all search terms are to be returned then the search terms should be linked with an ‘AND’ or ‘+’, e.g. diabetes AND medication. Conversely, if certain terms are to be excluded from the search then these can be preceded by a ‘NOT’, e.g. "diabetes research" AND NOT treatment.

Directed Searches

The search engine looks by default in all sections of the Case Study for the search term(s). It is possible to direct the search, however, and to look for terms in specific sections of the document. This is achieved by prefixing the search term by the name of the document section that is used in the document index. The valid section names are:

Index Field Name	Case Study Section
institution	Submitting Institution
UoA	Unit of Assessment
title	Title
summary	Summary of the Impact
research	Underpinning Research
references	References to the Research
impact	Details of the Impact
sources	Sources to Corroborate the Impact

So for example, to search for "astronomy" in the case study title we would enter: title:astronomy.
Note:

• The case of the index field name is significant, and should be lower case as in the table above.
• There should be no spaces between the colon (:) and the start of the first search term.
• The search terms are exact: e.g. plurals will not be found if a singular term is used.

Thus, to narrow down the “title:astronomy” search by looking for the term "Black Hole" in the section: Details of the Impact, the query would be modified to: title:astronomy AND impact:"Black Hole".
If the search is for two words in the same section then the field name must be specified for each.

Wildcard Searches

The search engine supports single and multiple character wildcard searches within single terms.

To perform a single character wildcard search use the "?" symbol, e.g. Latin?

To perform a multiple character wildcard search use the "*" symbol, e.g. genom*.

Note: You cannot use a * or ? symbol as the first character of a search.

Note: You cannot use a * or ? symbol within phrase queries using double quotations.

Tag Filtering

Although it is possible to include an institution’s name in the search query, it may not produce predictable results if the spelling or composition of the institution name does not precisely match the spelling used in the original document. The REF2014 search provides a collection of tag terms that are applied to each case study and which can be used in order to filter results.

Tag Type	Description
Submitting Institution	The higher education institution (HEI) that has submitted the case study. Note that in the case of joint submissions there will be multiple HEIs associated with a case study.
Unit of Assessment	The subject grouping to which the case study has been submitted.
Summary Impact Type	The main area impacted by the research.
Research Subject Area	Each case study is tagged with up to three research subject areas using the ANZSRC Field of Research Classification.
Impact Global Location	Cases studies may be tagged with one or more countries or cities identifed within the document text.
Impact UK Location	Tags for UK locations within the document text.
Interdisciplinary	If a case study is relevant to more than one major research subject area then it is classed as interdisciplinary.

The available tag values are displayed on the index tabs on the main REF2014 search page.

The Results page for each search displays the number of case studies associated with each tag value.

If you are experiencing difficult using the database, please email: researchpolicy@hefce.ac.uk