Springe direkt zu Inhalt

Finding a title, an abstract and keywords for a publication

Proper metadata is a prerequisite for good discoverability of publications. Metadata is structured data that contains information about other data, i.e. describes it in more detail. Such metadata naturally also includes the title, abstract and any keywords or subject headings assigned to a publication. As authors usually write or select these themselves, they have an influence on the quality of this metadata that should not be underestimated. Authors should therefore be careful when choosing titles, writing summaries and assigning keywords or subject headings to their publications. By doing so, they enable their publications to be found as easily and reliably as possible in catalogues and databases. Here are some tips on writing titles and summaries and selecting keywords.

Concise titles

The title of a publication is the most important metadata, as it is usually the first thing readers see. The title is the first point of reference for assessing the content and therefore the relevance of the publication. In search engine queries, as a minimum the title is always taken into account and in the search results, users will often decide whether to continue reading based on the title alone. A title is like an advertising slogan for the publication. Titles must therefore be concise and well considered. Here are some tips for choosing a title:

  • The title should be informative and precise, i.e. convey the most important information as concisely as possible – but not too briefly or superficially. Filler words and redundancies as well as self-explanatory expressions, such as "A study on ...", can often be omitted without any loss of information.
  • The title should already contain the important keywords.
  • Ideally, important terms should appear at the beginning and end of the title.
  • The title should not allow any ambiguities. These can often be avoided by simple rephrasing.
  • Avoid abbreviations in the title. (With the exception of very common abbreviations, such as DNA).
  • If the title is divided into a main title and a subtitle, the most important information should already be contained in the main title. Popular variations on dual titles, such as supposedly funny titles or quotes as titles, should be avoided and chosen only with care.
  • Consider any requirements of the journal or publisher, for example character limits and instructions on wording style. In the case of journals, also check previously published articles to see whether there may be an unwritten convention for the wording of titles.

Informative summaries

Publications almost always contain a summary of the content (commonly referred to as the abstract), i.e. a condensed version of the publication with the key aspects of the publication in a brief form. An abstract must be clear and easy to understand. The title and the abstract are the only metadata searched by many search services and the abstract is the first and often the only thing that readers actually read of a publication apart from the title. A well-written abstract invites readers to read the publication and engage with the underlying research. For abstracts, there are usually guidelines from the journal or publisher regarding maximum length and style. An abstract usually comprises 150 to 300 words. In the case of articles, the abstract often reflects the structure of the publication. Here are some tips for writing an abstract:

  • An abstract should only be finalised once the manuscript for publication is ready. (However, writing an abstract for a planned publication can be a good test of whether you are ready to publish).
  • Like the manuscript, an abstract usually requires several reviews and revisions. Allow time for this.
  • An abstract should not be so detailed that it requires citations, references or footnotes.
  • As with the title, filler words and redundancies as well as self-explanatory expressions should be omitted. Guides often recommend summarising the individual sections of the main text in a few sentences, stringing them together and then shortening and condensing them until only the essential content and information is retained.
  • Important keywords from the publication should be included in the abstract. This applies in particular if no separate keywords or subject headings are assigned.
  • In disciplines in which articles are structured along the four building blocks of introduction (with the research problem or motivation), methods, results and discussion (with the conclusions), the abstract should also follow this structure.
  • Read summaries of other publications (e.g. from the same journal) for inspiration and to familiarise yourself with common conventions.
  • As with the main text, it can be helpful to have the abstract read and checked by a colleague. People from outside the field can usually provide helpful feedback on the comprehensibility of an abstract.

Standardised keywords

When submitting a publication, authors can usually also assign a certain number of keywords or subject headings, which are included in the publication and metadata in addition to the title and abstract. In general, the number of keywords or subject headings is limited by the journal or publisher, but they can otherwise be freely assigned. And this is precisely where authors can increase the quality of the metadata by not just freely assigning keywords, but by choosing standardised keywords from a defined set of terms.

Such a predefined set of designations is referred to as a controlled vocabulary and is stored in special databases, for example in the form of so-called authority files. There are editorial specifications as to which items (terms, personal names, place names, etc.) are included in which spelling and preferred designation and to which alternative spellings and designations (synonyms) as well as foreign language equivalents these are linked. For example, different spellings of a city (Moscow, Moskva, Moskwa, Moskau) or the same subject matter with different names (economics, economic science) can be linked with each other; and people with identical names (Michael Müller, the politician, and Michael Müller, the sociologist) or homonyms (words with several meanings, such as a bank as a feature of a river or a bank as a financial institution) can be distinguished. In addition, relationships can be specified among them, such as music sub-genres or sub-areas of scientific disciplines. If designations from such a controlled vocabulary are used, they are referred to as subject headings, in contrast to freely assigned keywords.

The advantage of standardised subject headings can be illustrated using the example of a literature search. If no controlled vocabulary is used, search hits are only returned if the search terms are found in the metadata verbatim/literally – neither synonyms are found, nor can ambiguous terms be selectively filtered. If, on the other hand, a controlled vocabulary is used – which is usually the case with academic search platforms and library catalogues in particular – and corresponding subject headings are available in the metadata, significantly more precise and relevant search results can be returned.

Wherever possible, authors should therefore use standardised subject headings rather than free keywords. This allows for increased visibility and consequently also increased accessibility and citation of their publications. For example, the Integrated Authority File (Gemeinsame Normdatei, GND) can be used for German-language subject headings and the Library of Congress Subject Headings (LCSH) for English-language subject headings.

The following guide provides a good overview of how authors can go about this and how to search for standardised subject headings (so far only available in German): Vom Suchen und Finden: Handreichung zur Arbeit mit kontrollierten Vokabularen und Normdateien (2023, by Moritz Strickert, EVIFA Working Papers 1).

Related Topics