Matthew Knott

Something I came up with yesterday was a pretty neat AJAX-based tagging system. It's pretty and easy to use, and uses efficient javascript to create the tag graphics on the page.

 

Tags: Tags Content Organisation Tagging Management System Article

I was at a tri-authority collaboration meeting today, and had a chance to explain to people what I call the three-layer model for organising content, or the three-layer content organisational model to you and me.

Layer 1: Folders

At the top of the layer is the most rigid of structures, folders. Folders are great for organising things and keeping them tidy, but if you don't know off hand where a file is, you have to search, which is only useful if you know the document title or what it might contain. Folders are great for structuring information in a narrative sense such as documents > projects > regeneration project, but this relies on a well planned structure. 

Layer 2: Categories

A good bridging layer, categories, are very similar to folders, depending on how they are used. In fact it is very easy, especially in online content, to duplicate the functionality of folders but call it categories, using endless sub categories to contain documents. True categorisation should be high level. Where categories differ from folders is that they should be single level, and documents can exist in multiple categories, where as they can only exist in a single folder. Categorisation makes browsing to a documents location quicker, but there can be a larger volume of documents, so this can make pinpointing your required information harder.

Layer 3: Tags

Tagging is essentially structureless, categorisation on steroids. Tags represent both a documents content and it's meaning, implied or otherwise. There are hundreds of these little categories. Browsing would be difficult and time consuming, but search results are accurate and valuable. Tags should be added by a documents author, their tags represent what the document means to them, but tagging systems should allow others to add tags, building the documents repertoire for multiple perspectives. This variety of tags is what makes the search process more reliable and less linear.

Usage

Your documents should be floating objects, with the three layers all being implemented as views into your documents, this way you can provide the most complete user experience.

Tags: Tags Tag Document Organisation Methodology Model Categorisation Topology

I've just been dabbling with Apache's mod_rewrite functionality and it is super-sweet. It's one of those techniques that you just see everywhere these days.

I followed this example, which was easy to follow, to control the redirect.

I've implemented it on the tag system and also the categories and I'm really impressed with the results. Give it a try.

Tags: PHP Tags Blog Apache mod_rewrite .htaccess

I wanted to create a tag cloud, and looked at Cal Evans's Cloud Generator, which to me seemed overly complex, though very stable I'm sure. I just thought that it would be easier to work it out myself and tailor the solution to my needs.

It was actually easier to do than I thought.

1. I run a query against the top 20 tags, returning the tag name, and usage tally, ordered alphabetically.
2. I loop the results, adding them to two separate arrays, and calculate the maximum usage and minimum usage.
3. Now I have the maximum usage value, I calcululate each usage as a percentage of the maximum, which I use to give the tag a rank 1-5. 0% - 19% is rank 1, 20% - 39% is rank 2 and so on.
4. The tag is output as <span class="tag_rank_1"><a href="tagurl">Tag Name</a></span>
5. In the stylesheet I set the styles I want for each rank such as darkening and enlargement of the font.

Click here for the source code.

Tags: PHP Tutorial Tags Blog Cloud Stylesheet

Due to a failure in the heating system in work today, we were all sent home early. In fairness it was bloody freezing.

After a 3 hour power nap, I got up and finally completed the tagging system for the blog.

It works simply though an onBlur function on an input box, splitting the tags into an array and inserting them into the database. Each keyword is passed separately so that I can check if it already exists in the database, in increment the usage accordingly so that I can later create a tag list, sized around popularity.  

It's a start anyway, and it works well, which is all I want. Off on holiday tomorrow, and hoping to be off Wednesday to help Lisa.

Tags: Javascript PHP Cold Work Temperature Tags Blog Split