I’ve spent much of today collecting data for a chapter that I wasn’t sure whether I would end up including, but have decided to. I already had quite a bit but needed more. This is part of the result of today’s effort.
The chapter is about the visual and metadata languages of Perth. I’m analysing images from Flickr and Instagram which, based upon the inclusion of those images in groups (on Flickr), tags (on both platforms), and descriptions (on both platforms) are said to be of Perth. I’m also plotting network visualisations of the photographers and their networks/tags to see whether I can tell a story about the visual language of this city.
So far it’s proved really challenging, which is why I was thinking about not including the chapter. I feel like it needs a lot more attention to get to the point where it says what I want it to say. It’s the last chapter in my thesis, but I think it so nicely draws together the story that is told in the other chapters. I wasn’t originally planning on taking data from descriptions in Flickr, but I think I need to as it’s looking like Flickr users are less likely to use tags than users of other platforms, so it made my earlier data collection look quite sparse.
Something that’s been generally frustrating throughout my research has been the inconsistency of descriptors used by contributors to social platforms. I know that this is something I’m more aware of than most, given that it’s a core element of my study (i.e. the language of online storytelling), so no one else is to blame, although I think there is definitely space for better educating users on getting the most out of a platform’s affordances from a social/audience perspective.
My research has generally involved a lot of manual trawling through blogs, Twitter feeds, Instagram feeds, Foursquare check ins, Flickr profiles, and so on, to find the content that I know is there. It’s required many, many, many hours of following links to work out what is and is not suitable data for my research — and that in itself is something I do discuss in my results. Being a coding genius and knowing how to extract data from websites is only as good as the information that is written into sites, in text and tags. So much of what I’ve discovered has come from me being able to recognise a photograph of a place in Perth, or know that a certain restaurant is in Northbridge, and so on. It’s been challenging and awfully time consuming, but I think I’ve pieced together something that tells a nice story about how the people of Perth use social platforms.