Faculdade de Letras da Universidade do Porto - OCS, 15th INTERNATIONAL ISKO CONFERENCE

Font Size: 
Marit Kristine Ådland, Marianne Lykke

Last modified: 2018-01-31


The purpose of the paper is to explore tags and tagging on an information web site, as part of users’ information behaviour: how is the aboutness expressed, what facets are used, how is the meaning or topical content of tags related to the document content, what vocabulary is used by the taggers, and how are users’ opinions on tags.  We have chosen Cancer.dk as a case. This is an information web site that targets cancer patients and their relatives. We find it interesting to study tags in a system where metadata is not as visible to the users compared to many other systems with tags (ref).

More specifically, we seek to answer the following research questions:

  1. What characterizes tags on Cancer.dk?
  2. How are the tags on Cancer.dk compared to other tagging systems?
  3. What are the users’ views on and intentions with tags?
  4. What are the editors’ views on tags?
  5. Do the views of users and editors correspond with each other and with the nature of tags?


This is a user-oriented (Järvelin & Ingwersen, 2010) empirical, exploratory  case study (Bryman, 2012) where we studied users’ interaction with tags and tagging, and information behaviour in general on a certain web site, Cancer.dk. In a user-oriented study, the user is seen as a part of the system (Järvelin & Ingwersen, 2010). With tags, the user also provide metadata into the system, and thus user behaviour is crucial in understanding the whole system.

A combination of qualitative and quantitative methods provides a holistic view and thus a multifarious picture of the tagging behaviour at Cancer.dk (Bergman, 2008). The quantitative data, from log files, provides insight into what is going on and make it possible to explore whether an identified phenomenon is frequent or not. We counted tags of various types and compared the number of tags in different categories. We did not exclude any tag from analysis, The quantitative data proves an overview and identify some patters, and no tags were excluded from the analysis. The qualitative data provided explanations.

Tree studies were conducted:

1 Tagger study: Eight participants solved tasks using the tagging feature. They filled out questionnaires, and we interviewed them. The goal was to find out about their understanding, motivations, and opinions of tags and tagging, including their purposes for applying tags.

2 Editor study: We interviewed three editors about their experience and opinions on tags and tagging. The goal was to find out about their opinions, and to be able to compare this to the users’ opinions.

3 Tag study: A log from the tagging feature gave the largest data material. We analysed all tags that had been applied during a year. The data material include approximately 25 000 tags. We categorised the tags into various categories:

  1. Internal and external tags – indication on who applied the tag
  2. Lay or professional – do the tag belong to a lay or professional vocabulary
  3. Topical facets – what is the tag about
  4. Aboutness – the relationship between the aboutness of the tag and the aboutness of the article


Our analysis indicate a connection between computer skills and understanding of the tagging feature on one side, and a focus on applying tags as topical descriptors. Topical description is the dominating purpose when applying tags at Cancer.dk. Other purposes are also found: tags to explain the content, tags to evaluate articles, and tags to express wishes for additional information in articles to editors. These tags represent attempts to communicate with the system. All the participants agreed that topical tags are good, but they did not agree on whether other types of tags add value to Cancer.dk.

The different purposes users have when they apply tags is a challenge to the editors. They prefer tags that describe the topical content of articles. The tags studies challenge and contradicts the editors’ view on tags as mainly users’ voice or colloquial vocabulary into the system.

It is difficult to apply tags. Especially the aboutness categorization reveals challenges with how tags relate to the topical content of the article. Some tags are relevant subject descriptions of articles; many are not. They express users’ and patient’ experiences and needs, or seem to be misunderstandings of the tagging feature, or individual thoughts. Mixed together, the tags as a whole are difficult to utilize.

Internal and external taggers behaved differently. The internal taggers are from inside the organization behind Cancer.dk. It was easy to address them as a group and encourage them to apply tags. However, the crowd of external taggers was more stable. Their number of tags varied less during the logging period.

Tags from internal taggers cover more diverse categories and describe the article content from various angels. Their tags are more evenly distributed on tag facets, compared to external taggers. These results are in conflict with an expectation that external users can add new viewpoints to the systems. External taggers however applied more tags not related to the content of articles.

As an overall conclusion, we have found a disagreement between taggers and editors on what a good tag should be like. This explains some of the “wrong”, irrelevant, non-topical tags: They were intended, as taggers attemted to communicate with the system through tags as opposed to describe the content. Editors on the other side, saw topical descriptive tags as most useful, and found that the feature attracted too many tags that did not meet this criterion. These conclusions supplement other studies of tagging behaviour, where various metadata are more visible to the users when they apply tags.




Bergman, M. M. (2008). Advances in mixed methods research: theories and applications. Los Angeles: Sage Publications.

Bryman, A. (2012). Social research methods (4th ed.). Oxford: Oxford University Press.

Järvelin, K., & Ingwersen, P. (2010). User-Oriented and Cognitive Models of Information Retrieval. In Encyclopedia of Library and Information Sciences (Third edition, pp. 5521–5534).