This chapter reports a study on the tagging feature at Cancer.dk. Eight participants were interviewed after completing four tasks each. The tasks asked them to find information at Cancer.dk and to apply tags to articles when the information was found.
The participants liked the tagging feature and were all able to use it, even though some of them did not understand the complete functionality of the tagging system.
The participants retrieved articles and applied tags because they were asked to. Thus, their behaviour was not necessarily fully natural. But their reflections on what purposes tag have and can have for themselves and others are funded on their own experiences.
Our data indicate a connection between computer skills and understanding of the tagging feature on one side, and a focus on tags as topical descriptors. But other tag purposes are also found: tags to explain, tags to value articles and tags to express wishes for additional information in articles. Topical description is the dominating purpose when applying tags at Cancer.dk. All participants agree that a good tag has an understandable relation to the subject of the document it is applied to. With a focus on the use of tags, slightly different tag usage purposes appear: Exploring other taggers’ use of Cancer.dk, searching, re-finding articles and finding explanations.
Other findings reveal that taggers re-use words from the documents they tag or other information available at the moment of tagging (Bar-Ilan et al., 2010), but the participants do not reflect much about their sources for tag formulation.
All the participants agree that subject descriptive tags are good, but they do not agree on whether other types of tags add value to Cancer.dk. Especially tags that express wishes for information are controversial. One participant applied such tags. Other
participants had a strict opinion that such tags were useless and could even disturb the credibility of Cancer.dk. This is correct. The wishes from an individual tagger do not give any help in the dialog between users and Cancer.dk, and there is no reason to believe that the editors are able to identify the wish act upon it. These tags are misleading and not wanted in the system.
The focus on helping others may have been strengthened by the tagging feature as an extended narrow folksonomy. The feature does not support personal information management through tags. The helping focus is also visible in the applied tags. Many of them are synonyms or reformulations of article words.
9 EDITORS’ VIEW ON TAGS AND TAGGING
9.1 BACKGROUND AND PURPOSE
In December 2012, I interviewed the editors of Cancer.dk about the tagging feature and the tags. The feature had been present on Cancer.dk for abuot a year, so the editors now had experience with it. The tagging feature connects closely to Cancer.dk and is maintained by its editors. They write articles for the website in cooperation with medical and social experts, and structure the site. They edit and delete tags when needed. All together, these persons are responsible for the system of which tags are a part.
The end users’ view on tags and the tagging feature is the most important when one wants to find out how tags can be used on a site like Cancer.dk. However, the editors are the ones who maintain the site. They sum up the responses and the behaviour they receicve and observe, and change the site or individual articles if they find it necessary. They use experts and other recources available in the Danish Cancer Society when needed, and they decide what to do when they discover conflicting needs among the users. Their decisions frame the tagging feature and give it its shape.
Thus, they could give information about their thoughts and actions on tags.
I wanted to interview them to obtain their view on the tagging feature and the tags. I wanted to know their opinion about the tagging feature, good and bad tags, and any damage they believe the tags may do, if any. The editors did not access and analyse the transaction log or similar data, but they have a close relationship with Cancer.dk and its users. I wanted them to share this information.
9.2 METHOD
In addition to the editor in chief, there are two full time employees and one part time employee in the editorial board. I interviewed all three of them. To help them, there are decentralized Cancer.dk editors in all departments of the Danish Cancer Society.
They are responsible for content on parts of the site, but they also have other responsebilities. I did not contact any of them. It could have been interesting to have their view as well, but in this project is was a priority to talk to users and the editorial board. The users because they are the target group of the tagging feature, and the editorial board because they are in the crossroads between the users and Cancer.dk’
information policy.
Again, I chose interviews to let the editors give their own opinion. An interview guide contained questions to ask. See Appendix 2. It includes a set of questions that I asked all three editors. In addition, the interview guide has suggestions for follow up-questions and alternatives. I wanted the editors to comment on some of the results from the user study: Tagging purposes, use of tags and sources for tag formulation.
Thus, this information is listed in the interview guide for follow-up questions. I did not give the editors the information immediately in the main questions, because I wanted them to give their own view first.
The interviews were conducted in 2012, December 13-19, by telephone from Oslo.
The editors used phones at their desks at the Dansih Cancer Society in Copenhagen.
Methodically, these interviews were carried out in the same way as the interviews with the user test persons. But we did not meet in person, only by phone. This of course give limited possibility to observe non-spoken signals. On the other hand, these interviews were longer, and they were a follow-up on a long project that they had been a part of for over a year. The theme of the interviews was something the editors had a professional relationship to. Thus, I expected them to be able to describe their opinions with no particular restistance and that talking to them on the phone would give a sufficient imige of their opinions. This view was confirmed by the communication I had with both the editor in chief and the editors themselves before the interviews:
They seemed to be polite, but also honest and open about their view on tags and tagging. During the interviewes, the editors also included criticism of both the tags and the tagging feature. See chapter 9.3 for results. I see this as a second confirmation that this time, telephone interviews gave enough information about the editors’
opinions.
The tagging feature was funded partly by the Danish Cancer Society and partly through project funding from the research team. This may lead the editorial board to be more polite and positive than they would if they talked to researchers who were not involved in the tagging feature and who were not specifically involved with tagging at all. On the other hand, in this sense, the editors are also responsible for the site and its content. They did not take part in the tagging project to produce research results, but saw it as an opportunity to test tags on the site. When the logging period was over, they decided to take the tagging feature off the site. This was known to all of us before the inerviews. All together, there is reason to believe that the most polite solution for the editors was to give a balanced view on tags and tagging.
Again, the interviewer spoke Norwegian and the editors Danish. I discussed with editor E3 wether English was better, she was not very much used to Norwegian language. But we ended up with Danish and a Danish-influenced Norwegian instead.
No language problems were observed during this or the other two interviews, or afterwords when the interviews were transcribed and examined. In particular, I looked for misunderstandings because of different pronunciation or similar words that have different meaning. Because Danish numbers can be complicated to a Norwegian, I made sure during the interviews that I had heard what they had meant to say.
9.2.1 THE EDITORS
The three interviewed editors, all of them work together and know each other, together with the editor in chief, are the complete central editorial board behind Cancer.dk.
They are the ones that makes decisions and do what it takes to carry them out. See an overview in Table 32.
Gender
Female Female Female
Interview duration
34 minutes 41 minutes 47 minutes
Table 32 Interviewed editors
Only one of the editors (E1) had been in this position since the workshop in 2008. I met her there. The other two became a part of the editorial board later, and did not have the same inside knowledge about the tagging project and its history. Editor E3 worked part time for the Danish Cancer society, and is the one person who edit most tags. All together, the three editors had a different background different familiarity to the tagging project. This may incluence their viewpoints and feelings about the tagging feature, and thus the interviews.
The editors agreed on many things when it comes to tagging. The differences between them were mostly differences in what emphasis they put on diverse aspects on tags and the tagging feature. An example is to what degree they believe tags that they regard as bad will disturb users.
9.2.2 DATA COLLECTION AND ANALYSIS
I transcribed the interviews and read them thorough. As a starting point, the material was coded and categorized to identify viewpoints and characteristics on the tagging behaviour (Kvale & Brinkmann, 2009). The coding was mainly data driven, but
influenced by the pre-pepared questions the participants were asked. Thus, I coded and looked for their opinions about:
• Good and bad tags
• The tagging feature
• To apply tags
• Who should apply tags?
• Usage of tags
• Tag editing
I also read the transcribed interviews to find information related to findings in review, blog tag analysis, tag analysis, and tagger study. Thus, I coded and looked for:
• Tag facets
• Purposes when users apply tags
Some of the items above are closely related: Who should apply tags is related to what porposes these users have when tagging. And tag facets are related to what usage of tags is possible.