Software Engineering for Smart Data Analytics & Smart Data Analytics for Software Engineering

User Tools

Site Tools


This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision
Previous revision
teaching:theses:thesis-tags_and_order [2008/04/22 08:51]
teaching:theses:thesis-tags_and_order [2018/05/09 01:59] (current)
Line 1: Line 1:
 +====== Thesis: Tags and Order ======
 +This work investigates how tagging of information relates to predefined structured orders. ​
 +Large dynamic information occurs in many areas, e.g. organizations,​ encyclopedias,​ web forums or wikis and other systems that collect and maintain systematic information. Typically an initial ordering is predefined (e.g. a structure of wiki pages, or a taxonomy of concepts). While the amount of information grows, in most case the initial structure becomes inadequate and deteriorates (given the system is flexible enough like wikis are). Reworking the structure does help only for a short time, and gets more and more difficult with larger amount of information and more complex structuring.
 +Tagging information instead of sorting it into a predefined structure is a way to provide ordering in a very flexible manner. In contrast to structured orderings tagging allows to let a piece of information belong to multiple categories at the same time. Reordering in terms of changing the tags is different to hierarchical or otherwise structured orderings and seems to be more flexible.
 +Typical examples which show the limitations of predefined tree-like orders are: 
 + * how to sort photos? by date? by what's on the photos? by who is probably interested? by who took them?
 + * how to sort mails? by task?, by people? by date?, by relevance?, ...
 + * how sort bookmarks? by time? by topic? which kind of topic? ... 
 +Most things are more complex than trees, even trees ;-)
 +Interesting questions and tasks of this works can be:
 +  * how do tools support order? mail clients, photo manager, iTunes, amazon, ...
 +  * classify the diverse concepts of orderings (e.g. topic map, taxonomy, ontology, hierarchy, tagging etc.)
 +  * identify and analyze typical problems of large dynamic information spaces and give an  overview of solution possibilities for the above mentioned concepts of orderings
 +  * analyze the status quo of tagging. It is already widely used for user feedback e.g. on web site content and to support "​social computing/​networking"​
 +  * develop a prototypical framework for using information tagging. This framework should support the adaptation functions identified before and support using tagging. E.g. rename tags, unify tags, subdivide tags, find and access tags, define relation between tags?, etc.
 +    * the information tagging framework should be (at least conceptually) applicable for wikis, web sites, data bases for augmented reality etc.
 +    * use and test the framework with a small project on tagging for augmented reality
 +  * further questions could be: how does tagging differ from a hierarchical ordering when we order the tags? Is it useful to analyze the information pieces (e.g. cluster analysis, formal concept analysis to find new potential tags) in order to check or optimize adequateness of the current tagging?
 +The following sketches shall illustrate different orders:
 +====== Implementation Ideas ======
 +  * Implement a tag hierarchy browser. It should show a tree of tagged documents that is dynamically ordered based on user interaction. Folders are tags. I select a tag and all documents tagged with it become sons. Having selected the folder (tag) I select another tag, which then becomes a sub-folder and contains all documents which are tagged with both tags. And so on.
 +  * Imagine an operating system that organizes all items this way. You might have a predefined set of tags for each kind of item, say program, text document, web-link, image, sound file, etc. Now you can access anything in the order you need it. You can save tag paths for later quick access. Your system might update the search results for your saved tag paths in the background.
 +  * Imagine this web-based. You can upload anything and access it arbitrarily in multi ways in parallel.
 +  * Imagine this shared. We could simply start with web searches to retrieve the result set for a tag path.
 +  * Imagine this mobile and within a certain context. context can determine parts of your tag paths. This means you can focus you typical order automatically according to your current context.
 +  * Maybe we need more ''​intelligent''​ tags here, basically item constraints that can be used to restrict a given set of items.
 +====== Links and Literature ======
 +  * [[http://​​2008/​01/​08/​tagging-things-or-thinging-tags/​| Nice little observation about applying tags]]
 +  * [[http://​​reader|Google Feed Reader]] ​
 +    * allows to create a list of folders where you can sort your feeds into
 +    * allows for tagging each feed with the folder names as tags
 +If you are interested in this topic contact [[| Holger Mügge]].
teaching/theses/thesis-tags_and_order.txt · Last modified: 2018/05/09 01:59 (external edit)

SEWiki, © 2019