Ben Nadel
On User Experience (UX) Design, JavaScript, ColdFusion, Node.js, Life, and Love.
Ben Nadel at cf.Objective() 2014 (Bloomington, MN) with: Matthew Reinbold
Ben Nadel at cf.Objective() 2014 (Bloomington, MN) with: Matthew Reinbold@libel_vox )

Skin Spider : Auto-Tagging Downloaded Videos

By Ben Nadel on

To learn all about Skin Spider, click here.

To view the most updated application, click here.

To view the current code base, click here.

This is a small update but one that is very good for the usability of the system. I have added the ability to auto-tag videos as they are being downloaded. This removes a lot of manual work that the user would have to do in order to categorize the downloaded videos.

In order to accomplish this, I only had to make a few small changes. First, I updated the XML database schema in the ColdFusion configuration bean, Config.cfc. I added the column "keywords" to the tag.xml data file configuration. This column will hold the keywords that are used to auto tag videos.

Then, I updated the edit_tag.cfm ColdFusion template. I merely added a new form field and the code that commits it to the database. That took about 2 minutes. The form field is an (X)HTML TextArea. It is designed such that each keyword is on a single line. This allows the user to enter multiple keywords per line as a phrase. I thought about using quotes to denote phrases, but that just increases the complexity of what we are doing in a highly unnecessary way. Doing it this way allows us to use the ascii characters 13 and 10 to be used at list delimiters that naturally break up the keywords whether they are a single word or a group of words.

Then, the final step was to update the spider_video.cfm ColdFusion template. Again, this was a fairly easy task. I get the gallery name of the video that is currently being spidered. If you can remember, I pass in a gallery ID to the spider_video.cfm page to help the template figure out which video to spider. Anyway, I get the gallery name based on that passed in ID. Then, once I have downloaded the video, I loop over every keyword and match it against the title using a FindNoCase(). For any keyword that matches the gallery title, I automatically tag that video with the current tag.

It's nice. I think this very small update will make a big difference in the user's experience of the application.

What I have noticed as I test this feature is that many of galleries have really poor titles. This will lead to bad auto-tagging. I think that as I get more into it, I will have to figure out more ways to gather information about the video and gallery. Perhaps the meta descriptions and keywords, although, I am not sure the this type of web page uses those effectively either. We shall see. Baby steps to the ultimate application.

Reader Comments