There are several ways of searching Dimensions: using keywords or phrases; searching for DOIs or other document identifiers; by entering an abstract or longer piece of text to find similar documents; using the advanced search to build complex queries using automatically suggested related concepts; or by starting a search from a filter instead.
Overview
Keyword search
Searching with keywords is easily done in Dimensions, simply by typing one or more terms or phrases into the text box at the top of the page, and then pressing ENTER. This will then search for any document which includes all of the keywords entered.
Search indices and options
Keyword or text searches can be run in two different search indices:
Full data search
The search terms or phrases will be searched for in all metadata fields that are available for that document (for example, for publications, this may include affiliation, journal name, authors, etc. as well as the title and abstract). In cases where we have also indexed the full text (roughly 70% of the publications in Dimensions) this will also be searched, even though the full text may not be available to see within Dimensions itself. This means that many results that are returned may not include any obvious mention of the queried term(s) on the document's detail page. This also applies to searches within the patents data, where many full text documents are available for both searching and reading.
Title and abstract search
The search will be restricted to the title, subtitle and abstract fields only, when searching in publications or grants. When searching in clinical trials and patents, the title and description/summary will be searched. In addition, for publications, the following fields will be searched as well:
- Various unique document identifiers (Dimensions publication ID, DOI, Crossref ID, PubMed ID, PubMed Central ID, ReadCube ID);
- MeSH Terms (for publications indexed in MEDLINE);
- (Author) keywords (provided by some publishers and for publications indexed in MEDLINE).
DOI search
Within publications, you can also run a search using the Digital Object Identifier (DOI) of a document. Simply select "DOI" from the options in the search box, and then paste the DOI of interest, for example, 10.3389/frma.2018.00023 (DOI URLs may also be used), and press ENTER.
Entering the stem or prefix of a DOI for a certain publisher or journal can also be used in this way in order to retrieve publications from a certain journal or publisher (although this can also be done using the "Source title" or "Publisher" filters on the left-hand side). For example, a search for the DOI 10.3389/frma.2018.00023 will return the individual publication (as will a search for https://doi.org/10.3389/frma.2018.00023), whereas a search for the stem, followed by an asterisk, e.g. 10.3389/frma*, will return all publications with the DOI stem for the journal.
You can also search for more than one DOI at a time by simply pasting multiple DOIs into the search box (or concatenating them using Boolean OR). Note that querying for large batches of identifiers is not currently supported. While 300-400 DOIs may work fine, a larger number may result in a server error. Adding an option to query for several thousand identifiers is on our roadmap for a future release of Dimensions.
Notes on searching using keywords
Singular and plural
By default, Dimensions will return results including both the singular and plural forms of your search terms, i.e. searches for bird, birds, or bird OR birds will all return the same results. This is irrespective of whether you use phrasal search / quotation marks around the terms.
Spelling variations
Both American and British English spelling exist in Dimensions. In most cases, searching for one variant will also bring back results for the other; for example, a search for aging will also return results including 'ageing'. If you should come across examples where this isn't the case, please let us know via the 'Send Feedback' option in the 'Support' menu at the top.
Case
Boolean operators (see below) need to be capitalized; however, for other words, case is ignored, and the search will return the same results, whether or not letters of a keyword are in upper or lower case.
Punctuation marks and special characters
Depending on the Dimensions product variant you're using, some special characters or punctuation marks, such as "~", "^", "?", ":", "*", or ".", may be ignored or removed (for example, covid19~1 is interpreted as covid191) by the search, or have a special function (see "term modifiers" below). Also, leading wildcards ("?" or "*" at the beginning of a term) are removed and will be ignored.
The following characters are escaped (replaced by a whitespace) if the term or phrase is not enclosed by quotation marks: "+", "-", "&", "|", "!", "{", "}", "[", "]", "\", "/", and the non-breaking space.
If your results don't match your expectations, review your query for punctuation marks or special characters and make sure you're using them correctly.
Boolean operators
Keywords can also be combined using boolean operators, and these must be entered in ALLCAPS.
| Operator | How to use |
|---|---|
| OR | If either term is found then the item will be returned. For example, green OR blue will return any item that has either 'green' or 'blue' within it. |
| AND | Both terms must be present, so a query of green AND blue will only return items where both terms are found within a single item. |
| NOT | Excludes items that include this term. So green NOT blue means all items with 'green' in them will be returned unless they also have the word 'blue' as well. |
Please note that NOR, XOR, and NAND are not supported.
Using parentheses
Parentheses should always be added when using boolean operators to ensure boolean logic is applied as expected. For example, if you enter (green AND blue) OR purple, Dimensions will return any items that have both 'green' and 'blue' plus any items that include 'purple'. But if you enter green AND (blue OR purple) then it will return any item that has both 'green' and 'blue' or 'green' and 'purple'.
Term modifiers
Term modifiers can be used in combination with keywords and / or boolean operators to create more complex queries. The following modifiers are available:
| Modifier | Name | How to use | Available in... |
|---|---|---|---|
| " " | Quotation marks | If you put text within quotation marks, Dimensions will find an exact match only. "green blue" will return only items which include that exact phrase. Items which include 'blue green', or where both words are found in separate locations, will not be retrieved. | All Dimensions versions |
| ? | Question mark | The question mark can be used as a wildcard for single characters. For example, the search string te?t would match both test and text. Please note: a question mark within a search phrase (using quotation marks) will be ignored. | Dimensions Plus and Analytics only |
| * | Asterisk | The asterisk is a very useful feature that allows you to look for word stems as it is used as a wildcard for multiple characters. If you enter green*, Dimensions will look for any words that start with 'green', such as 'greenery', 'greener', 'Greensmith', etc. Be careful with this, as ill considered use may return grants for unexpected reasons. For example, a search of plas* hoping to find 'plastic', 'plastics' 'plasticity', etc., will also bring back 'plasmodium', etc. Please note: an asterisk within a search phrase (using quotation marks) will be ignored. | Dimensions Plus and Analytics only |
| " "~ | Proximity search | This looks for terms that are within a specific distance from one another, with 'distance' referring to the number of term movements (i.e. number of words in-between as well as word order) needed to match the specified phrase. For example "apple pie delicious"~3 will search for the words 'apple', 'pie' and 'delicious' with a maximum distance of 3 term movements:
| Dimensions Plus and Analytics only |
Similar documents search*
*Available in Dimensions Plus and Analytics only
Using this method (formerly called "abstract search"), you can also search for documents which are similar to an abstract or other longer piece of text. Simply select "Similar Documents" on the right side of the search box, paste the text into the search box, and then press ENTER. Dimensions will then extract relevant concepts from the text and return the most similar documents. The similar documents search can be used with all available content types. It is therefore possible to request grants similar to a specific publication abstract, etc.
Abstracts of methods papers or highly technical wording will be likely to bring back a smaller number of results which are highly related in scope to the entered abstract. Using a review paper abstract or less technical piece of text is likely to lead to a greater number of more general results.
This type of search can be extremely useful when trying to identify publications or grant funding based on a specific topic which would otherwise be hard to define through complex keyword searches alone.
Text used in the "Similar Documents" search remains private to you and is neither uploaded nor stored in any way in the Dimensions database.
Changing the number of results and applying filters
The search for similar documents relies on complex weighted combinations of automatically identified noun phrases to identify highly similar documents, and due to the complex processing required for this, the results shown for this kind of search are limited.
By default, Dimensions returns up to 500 most similar documents. You can change this threshold manually to 1,000 or 2,000 documents using the dropdown at the top of the search results list. When applying additional filters, such as a research organization or a year range, Dimensions will recalculate the results and return a new set of similar documents up to the selected threshold - but this time within the body of documents meeting the criteria of the search and applied filters.
Advanced search**
**Available in Dimensions Analytics only
Selecting the "Advanced" option in the search box allows you to build more complex searches, either by searching within specific metadata fields, asking Dimensions to suggest terms and concepts co-occurring with the keywords you're looking for, or searching for a set of up to 25,000 identifiers. Read more about using the advanced search in Dimensions.
Starting a search from a filter
Searches do not require the use of keywords or entered text of any kind, and can instead be directly undertaken using the filters on the left-hand side. For example, a combination of filters can quickly and easily produce results for searches such as "all open access papers classified as 'Artificial intelligence and image processing' published in 2017, with at least one author affiliation with University College London". (Not all filters are available in all Dimensions versions).
