Ideally, the finished product should be able to take information from any source. However, in order to limit difficulty and only have one thing to worry about, the project will be focusing on Twitter. Twitter is almost perfect for this project because it has a huge volume of publicly accessible data and a well documented API. The front end of the program will be called Twemotion, but the project’s modular design should allow this to be replaced with any front end, along with the associated data source.

So in order to forge ahead with implementation of the project, a solid delivery plan needs to be laid down. There are a variety of forms that the end product could take: stand alone application, installable doohickey, web application, and so on and so forth. Each option has its own set of merits and drawbacks. Anything that must be downloaded and installed is extra effort on the part of an end user – extra effort which some may not be willing to expend based on how much (or little) interest they have in the program. Additionally, there is an element of trust required to download and install things from a strange man. These factors are not so much of an issue with my project; after all, I am only trying to impress a select few people, and they’re probably going to go through whatever ridiculous measures I have in place for the sake of being able to mark the project.

None the less, for ease of delivery and demonstration, I have elected to produce a web application. Ideally, this will allow me to develop bits and bobs of the program entirely independently of the computer I am working from. Work goes ahead with looking at assorted server-side scripting language libraries for interacting with the Twitter application programming interface (API). At the moment I’m mostly working with servlets/jsp, but that may change if I find something easier. After all, reimplementation in another language shouldn’t be too difficult once the groundwork is laid down.

And here is the rough diagram of how it will all work:

That’s all for now.

The more I continue my research, the more obvious it becomes that my work will heavily rely on a dictionary of words that have some pre-described mood (and possibly strength) associated with them. As a reminder to myself of the kinds of words that must be recognised, I have coded a small program that allows a user to add words to a list. Pretty much just an interface to a text document that stops one from adding duplicates. Nothing too flashy. I found an interesting list to start from here. I couldn’t locate a contact address, so I hope the author of http://copywritingkid.com/?p=17 doesn’t mind me lifting the list of words to use as a base. Further than this, I found a rather useful looking list of words that have been tagged, but permission is required to obtain the software (maybe even the list?) so I won’t mention that until said permission is obtained.

More to come after I’ve slept.

For one of my second year modules, I had to write an essay. My chosen subject was the nature of morality in virtual reality, and to what extent our societal morals should carry over. Here it is.

I made a calendar template by drawing one out on an A4 piece of paper, scanning it, and tidying it up. I’m quite happy with how it came out – easy to duplicate on a large scale, but maintaining a hand crafted look.