For my thesis I have been adventuring into the abyss of sentiment analysis on Twitter for corporate entities. Most sentiment analysis has been focused on politics and political movements, there has been some great research about the Arab Spring and other major political movements. Sentiment analysis has little research about how corporate entities are viewed or engaged with. This type of research is usually proprietary and kept in-house for marketing and outreach purposes. There are some obvious cases of prominent anti-corporate sentiment, see the Pepsi-Kendall Jenner disaster, but mostly corporate sentiment is a bit invisible.
A large component of my dissertation is to manually classify tweets into three categories; neutral, positive, or negative. The process of classifying 4000 tweets is a bit overwhelming and time consuming, but also an interesting experiment in language.
It has brought forward a series of quite a philosophical questions.
What is neutral?
What is negative?
What is positive?
I’ve had to think very critically about the language I associate with positivity, negativity, and neutrality. How we use language says so much about who we are. There’s the old adage “we judge ourselves on our intentions and others on their actions” and I think that is definitely well represented in Twitter. I try to be conscious before I put an opinion out in the universe and it has been interesting to see that that is not a universal perspective.
Expressing frustration at poor service or being ignored is not something to which I am immune. When my internet was being installed I definitely expressed my frustration on Twitter.
I don’t feel guilty for expressing this frustration, but it was a bit pointless. It wasn’t an outcome oriented action. I didn’t call their customer support to find out where my tech was, I didn’t provide real feedback. I just complained. Of course, complaining is cathartic and can relieve stress. I just don’t know how the person on the other of the tweet felt, knowing that I wasn’t really interested in a response but just in being a bit miserable at them.
I’ve still got quite a few of those four thousand tweets to work through and maybe I’ll change my perspective again. For now, the minutiae of reading individual tweets for sentiment is a little adventure into other people’s worlds.
If someone walked up to me and asked me to account for my time for the next six months I would be horrified. Not just because a strange human had spoken to me but because I find that level of planning delightful and terrifying.
The way my brain works, for better and for worse, means that some things “Definitely must be planned.” and other things “Absolutely can be changed.”
When I get an idea in my head I tend to do whatever possible to make that happen. A prime example is the 40 or so hours between my thinking about shaving my head and my doing so. Not the worst choice I’ve made, but the point remains the same.
Accounting for a somewhat ineffectual and often erratic brain can be really challenging. Making a work plan for something like a masters thesis allows you to account for lots of things, like illness or vacation. But, how do you build in time for other, less visible hurdles?
Like most things in life, there isn’t really an equation for invisible illnesses.
Assume you will spend X weeks doing mania induced work that will be illegible and wildly off topic.
Add to that the X number of days you will be unable to leave your house or contemplate anything beyond your own insignificance.
It’s also a bit defeating to plan for your own highs and lows. What is the difference between experience based planning and catastrophizing? Unfortunately, I don’t think this is an answerable question. I think all I can do is try to make the best possible plan and purchase a large fuzzy blanket to live in. Perhaps some chocolate.
I have spent the last week staring at my completed independent study assignment.
I’ve never been good at letting go of things. I have special movie ticket stubs for the last ten years in my wallet that I carry everywhere. I have library cards for the last five cities I’ve lived in. I have discount cards for companies that don’t exist in this country or even in the last place I lived.
Even though I spent an inordinate amount of time making unnecessary pie charts that didn’t make it into my final version, I still enjoyed writing this assignment.
Letting go of my independent study was really hard. It was a way of saying that I was finished with a topic that is very close to my heart. I’m sure I will continue to learn and read about altmetrics, not just in a professional setting. I hope I can make my tiny square of the universe more inclusive, diverse, and valuable.
I adore formatting. Deciding how to structure paragraphs? Check. Justifying all the text so it looks right? Check. Figuring out how to embed tables, a glossary, and charts? Check. Making sure font size looks correct for footnotes? Check. Doing all of this while listening to Humble, Kendrick Lamar’s new single? Check.
There was one main aspect of formatting that I had not expected. It look me a few weeks of going to university in London to realise that A4 was the standard paper size here. This is somewhat a shameful admission. I really thought it was more likely that one of my lecturers had an odd preference than there be a different standard here.
Despite having this information I still manage to forget about the existence of A4. Before printing a copy of my assignment to do final edits I had to figure out how to change my Google Docs settings to A4. This wasn’t terribly challenging, just curious that even though I’d lived in the UK for 18 months I am still very resistant to A4. It is always interesting what standards you can’t shake from where you grew up. For me, A4 will always seem weirdly long, as I imagine Canadian paper looks short and stubby to Brits.
I know this is likely a meaningless statement that impacts no aspect of the universe but, I hate A4. 8 1/2×11 forever.
Creating a glossary can be a tremendously frustrating activity. What seems self explanatory to one person can be completely alien to another. There is always a moment when writing a piece of work, academic or otherwise, where you need to decide if something needs to be defined. “Does the reader need additional information? If so, am I obligated to provide them with this information?”
Language is such an important part of how we communicate and making assumptions can cause a higher barrier for entry for the less familiar.
In recent months I’ve fallen into the term “explanatory comma”. The idea is that there are some coded words or concepts that might need to be explained to a wider audience. The term is often used in the context of race and culture when discussing questions such as, “Who is responsible for educating the dominant culture?” “Should people seek out additional information themselves, or is there a requirement for minority groups to provide annotation in their discussions?”
While the explanatory comma fulfils a different function in academia, it does still have its place. How much can you assume about the reader? Is anyone entitled to information? Or should we be encouraging readers to fill in their own gaps and further their own knowledge?
There doesn’t seem to be a right answer. In explaining too much you risk annoying an experienced reader. In not giving enough you risk alienating someone new the to field. Like so many other things, defining the “obvious” seems to fall to emotion. For all intents and purposes, “What feels right?”
For further information about the idea of the “explanatory comma” I would strongly recommend this episode of Code Switch that talks about the fallout from an episode about the legacy of rapper, actor, and activist Tupac Shakur. http://one.npr.org/?sharedMediaId=504482252:505487421
In tandem with my independent study, I have also been working on an assignment about resource collection and presentation. The assignment entailed creating a document of annotated resources about a topic and presenting it in a valuable way to a specific user ground. Despite being a non-traditional subject, I chose hip-hop for beginners.
I love hip-hop and always have. The range from party music, lyrical poetry, and gangsterism means I was never short of stories that were interesting for me and universes to investigate.
In the last few years I’ve made the important discovery of the wide variety of items that could be considered documents. The word document previously conjured up in my imagination dusty archives and filing cabinets. Doing this project has been a great way to harness this new found perspective on “What is a document?” CDs, flyers for shows, posters, and more are all part of the rich documentation associated with hip-hop.
There is a real disparity in the way documents from different fields are treated. Fields that are considered non-academic are less likely to have traditional archives. I was very excited to find two discrete archives containing documents related to hip-hop music and culture. Despite my excitement I was also concerned about the limitations of having these materials in only two locations, relatively close together on the East coast of the United States.
How we treat the cataloguing of certain types of history exposes our perspective on their value. Is a flyer for an N.W.A concert any less important than a program from a performance at La Scala?
Re-listening to some of my favourite albums has been a wonderful experience and helped remind me of the first time I heard those songs. I hope my project will of value to others and to help spread my deep belief in the important of hip-hop. More than that, I’ve reminded myself that hip-hop is a key part of who I am.
I am as much the Queen of the Night Aria as I am To Pimp a Butterfly.
Making charts, not unlike my schedule and list making, is often just a filler task that makes me feel that I’m being productive. Producing a chart is dangerously easy in Excel or Google Sheets. Some light copying and pasting gives me the undeserved emotional reward of having done some work. It also helps that I like my charts to be colourful, so it’s essentially a fun craft.
Remarkably, this time around, I’ve managed to surface some really interesting stories from my data. I don’t know if I would have been able to extract these ideas from a spreadsheet of numbers. Seeing how the numbers related to each other allowed me to understand the relationship between different countries’ data.
So much of what is interesting and valuable about altmetrics are these qualitative stories and being able to represent the data in a meaningful way. If you look at the above chart you can see that Rwanda and Belgium have proportionally very similar instances of being cited in policy documents tracked by Altmetric. Further, Nigeria seems to have a very high rate of being cited in policy documents.
I was very surprised to see the very strong Canadian representation in regards to publicly available post-publication peer reviews. Oxford University, the university chosen to represent the United Kingdom in this sample, was much less visible than expected. I had expected the United Kingdom to dominate across all types of altmetrics.
As mentioned previously, these insights are something that I wouldn’t necessarily have been able to contextualise in looking at just the raw numbers in a spreadsheet. The lessons seems to be that visualisations, like most things, are helpful in moderation.
Finally, I’m terribly delighted that I got to give this blog post a silly pun title.