The objective is to get better search results when searching through large quantities of data – for use by other users or as a reference source for oneself.
Tagging is the allocation of catchwords to elements that will be found again when these words are used. One can regard it as a type of classification, like a glossary – but more flexible.The classification of information began long before the computer age. In the specialist filed of typography, we know the classification of scripts. Whereby scripts are classified into groups according to individual characteristics like serif or the degree of curvature. Every script belongs to a group or class and other scripts within the same class have the same characteristics – and look fairly similar.
The conventional method to structure data in computer systems is to file it into hierarchical folders. The folders and sub folders as a rule, usually have distinct names built on one another. Data is then classified/identified according to its characteristics and filed in a folder – and only in his one folder. If a customer sends me a logo during a project, I could file this logo under: customer/project/graphic/logo. If however, this logo will be used for more projects, it would be better to open up a more generic file: customer/logo.
Through tagging, unlike with hierarchical folder structures, more classifications can be given. The logo could have the tag – customer, project X and project Y, so that the logo can be found under the customer as well as under the different projects. The prerequisite for this type of data filing is of course, a data system or database that supports tagging.
Assining relevant catchwords
With tagging, the question is often which catchwords should be given and by whom. If an open system is used, everyone can enter any tag. If tags are provided to groups, then different people could interpret almost identical tags in a totally different way. Another difficulty is that deviating word forms occur, through typing error, inflected words or other language variations. Difficulties can also arise if tag names do not make sense, if they are too specific or mean something to only ne person. These complications with open tagging can lead to imprecise searches and to no hits.
For large groups it is evident the quality of tags used improves. When larger collections of tags are available, users tend to use existing tags, so a structure organizes itself. A good example of effective social tagging is the bookmark organization of del.icio.us.
Administrative Tagging
A method we use to keep proliferation under control is administered tagging. With this, the tags used are determined by ‘Admin’ and for example can be chosen from a list. When new tags are added, care should be given to ensure that overlapping with existing tags does not occur and that a uniform spelling is used.
Another opportunity with administration tagging is the introduction of different levels. There can be a term associated with a value. For example quality can be described as good or bad. Without association ‘good’ would not be a useful tag. In this way elements can be found where quality = good have been tagged,
There is always more and more complex information that needs to be administered, structured and retrieved. Tagging helps us handle large amounts of data.
