Having just completed a subject devoted to Describing and analysing information resources as part of my studies, I can now look back at my initial understanding and chuckle. It was limited to MARC and Dewey, an idea of subject headings, authority lists and a very vague idea of Z39.50. Oh dear, how little my knowledge base was….
Whilst, there is still a great deal to know in this field, concerns about what is relevant seem unchanged since Younger (1997). Younger asserts that the basis of cataloguing is still the same, in that resource creators want their resources to be found and information seekers want to be able to find resources (p.463). Furthermore, it is clear that a uniform approach is not an easy achievement, nor even a desirable one. Information agencies have different requirements in terms of resource description. Hence “One size does not fit all” and a focus on interoperability between standards becomes the imminent aim (p. 465).
This post, however, will not be focussed on the different standards. The weblinks embedded in my blog provide examples of the standards. Instead, the focus is the benefits of keywords or hashtags, as provided by ordinary people, in the facilitation of information retrieval.
Guy and Tonkin (2006) define folksonomy as “a type of distributed classification system…. usually created by a group of individuals, typically the resource users. Users add tags to online items, such as images, videos, bookmarks and text. These tags are then shared and sometimes refined.” Furthermore, hashtags are a useful addition to formal classification though not a replacement. Moulaison (2008) points out that the application of tags is defined by the purpose of the tag and ranges from personal retrieval needs with exo-tagging to promotion of user generated content with endo-tagging (p. 101).
In assessing the usefulness of tagging for web retrieval I have chosen the mobile photo sharing application Instagram (IG) that can be viewed online via web.stagram or with statigram. The experience of tagging in IG corresponds to Moulaison’s (2008) concept of endo-tagging user-generated content in the form of photographs that are uploaded and shared with the larger IG community internationally.
Moulaison (2008) suggests that user generated content that is uploaded such as videos or with photo sharing apps the tags can be “more stable and long term”. (p. 107). This is definitely the case with IG where the app has a search function embedded. You can search by users or tags to find either contributors or actual photos in defined categories. The result of the tags, and this search function in IG, is a system where the majority of contributors apply mostly accurate tags that benefit all. Beyond this, the tags applied are endo-tags by the contributor of the photo and they can be modified or added to at any point. This ability to modify at any point is great because you can correct misspellings. Other users will often pop in to comment feeds and provide suggested tags that may have been missed or the contributor may not have known, for example an artists name of the artwork photographed. This tag is often then applied enabling this photo to be located in this category or location.
Critiques of folksonomy include the perception that they are imprecise, inexact and too personal (Guy & Tonkin, 2006). Moulaison (2008) citing Tennis (2006) also mentions the idea that social tagging is for personal benefit (p. 108) as a type of filing system where one can locate resources quickly at a later time. In IG elements of personal benefit are evident with the more obscure rogue tags assigned that only one person contributes their photos to. This is evident with some travel photos shared for instance where the tags incorporate the name of the person contributing the photo and place eg #sarahinparis.
In my photo-sharing experience with IG, the tags assigned are often agreed tags by the larger community. However, as there is no limit to the amount of tags you apply, you essentially can incorporate a mixture of tags that are both for a broader benefit, where everyone can search and locate the photo with other similar resources, to more personal tags that an individual can search to locate their own personal category where the photos are only contributed by that person and pertain to something specific. An example may include a series of photos that act as a photo essay and grouping them together with a specific hashtag such as #photoessayeiffeltower2012. However, as Guy & Tonkin (2006) point out the use of single purpose tags are not the dominant. As the purpose of IG is to share to a wider audience the consensus on shared meaning becomes the norm.
Another critique of folksonomies is the argument by Shirky as cited in Guy and Tonkin (2006) of the lack of synonymous relations. In IG the manner in which some users have tagged photos suggests the idea of synonymous relationships as present. As an example, #wheatpaste, #pasteup #poster provide the relations between these three words as being synonymous with each other and act as descriptors of the same type of art within the broader category of #streetart and #graffiti, two tags that are also often applied as synonymous categories. Users often apply as many synonyms as possible thus enabling the photo to be located in each of the tagged areas in IG. This is very useful when searching as a person may only be aware of one term and yet still be able to locate this resource because the different terms have been tagged. Beyond this, as an international app, often the tags from contributors in countries where English is not the dominant language will have tags ascribed in both English and the language of the home country too. Potentially multilingual descriptors could be applied.
Tags in IG demonstrate a type of hierarchy with broader categories and narrower categories. They enable relationships to be defined between terms for example defining the photo as streetart (what it is about) tagging what type of street art such as “stencil” or “pasteup”, tagging the artist name of the streetart (creator) and tagging the location (geography) from broader to narrower #Australia #melbourne #brunswick #laneways #zombiedancelane and so on. Effectively, tags supply metadata and the more tags supplied in IG the more specific is the description.
Some tags also suggest what equipment the photo was shot with eg #iphone #iphoneography #nikon300, the shutter speed and may indicate if the photograph has been modified with filters or even with apps. The interrelations of users can be indicated via tags too. For example there are IG meetups organised, where groups do a photo walk and then post the results to the tag of the #instameetup with place. Another example is #mobilephotogroup of photographers dedicated to the use of mobile phones in photography. People share photos with each other by making them available to specific hashtags dedicated to the purpose of taking and modifying or altering the original and then reposting the result. They organise exhibitions and have dedicated hashtags that repost the happening or attendance.
In resource description the focus has been on controlled vocabularies as a process of increasing uniformity in description and disambiguating terms. Consequently, the use of natural vocabulary or folksonomies is perceived as problematic. A point worth mentioning, is the suggested solution by Hider (2008 p. 183-184) of the imposition of a selection of controlled vocabulary terms that can be chosen by users for tags. However, having participated in a number of photography sharing apps including StreamZoo, PicYou, Tadaa, EyeEm and mobli, the apps that have provided a designated vocabulary list to choose from and, imposed limitations on how many tags to assign, produce description results that end up being too broad and less specific.
As an example, EyeEm only allows the assigning of one category for the type of photograph and the assigning of location. It does not allow further elaboration except what you may include as a title for the photo. If I took a photograph of street art that is a Paste-up of a deer by the artist Kaff-eine in a particular laneway at Fitzroy, Melbourne, Australia, with EyeEm all I can assign is “Streetart” “Fitzroy”, “Melbourne”. This is very limiting and not specific enough. Whilst the resource is placed in the broader category of Streetart and the location, these are the only two areas I can search to locate the work.
The audience base and user generated content of the other photo sharing apps appears significantly smaller than IG thereby it is still easy to locate the specific resource. However, on IG, Streetart, as a broad term, produces a result return of over 500000 photos. Consequently, the need to search more specific terms becomes paramount. In EyeEm there is no ability to search for the artist or the specific type of art either, which may produce a more accurate and specific result.
It is very useful to be able to search by artist for a streetart piece as well as by sub genre and even for far more specific locations. This is the advantage of applying more tags; as the description becomes more specific and enables the resource to exist with other resources that are also identified as having that specific category too.
This just touches on some aspects of how tags are utilised within IG and some of the purposes that tags demonstrate. The benefits of tags in web environments are certainly evident and not just related to a personal filing system. I suspect like other types of classification systems tagging or folksonomies will continue to evolve with purpose and function. Tags in IG do indicate an ability to effectively apply descriptions and can increase in specificity of description because there is no imposed limit on how many tags can be applied. The ability to title work or even contribute a mini or longer write-up about a photo further contributes to the description.
Guy, M & Tonkin, M. (2006) Folksonomies: Tidying up tags? D-Lib Magazine, 12(1)
Hider, P. (with Harvey, R.) (2008). Organising knowledge in a global society. (Rev. ed.). Wagga Wagga, NSW: Centre for Information Studies, Charles Sturt University.
Moulaison, H.L (2008) Social Tagging in the Web 2.0 Environment: Author vs. User Tagging, Journal of Library Metadata, 8:2, 101-111.
Younger, J.A. (1997). Resources description in the digital age. Library Trends , 45(3): 462-481