Post 6: Data Scraping

With the topic of Asylum Seekers and Refugees being so deeply entrenched within today’s societal context, these research methods have been crucial in furthering my own understanding of the issues intertwining relationship with the community. This latest task has allowed me to further understand this issue on a community and public level by engaging more deeply with social media trends and posts. By doing so, I have been able to engage with a predominant source of primary communication in today’s society.

I chose to scrape the social media site Twitter. Twitter is often cited in the news in showing the public’s opinion in short bursts (140 characters) and allows for an overall general feeling for attitudes based around the issue. As well as this, with many prominent news sites and politicians on the platform, it means theoretically we can gage trends and patterns in public opinion in accordance to fluctuating opinions and policies. Saying that, in order to retrieve quantitive data that I would be able to draw questions/solutions from, it was vital to consider the definitions that were being used in the search. Initially, using the technique shown to us in week 5, I started out very broadly searching the terms “boat people”, “asylum seeker” and “refugees” to see if I could try and find any links to the dates in which these terms were used the most (eg data leaks, international events). However, this only served to give very broad search results and therefore I had to refine my search terms in order to filter through the data I could find on twitter.

I redefined what I was trying to engage with in this particular topic: the way the media portrays the refugee crisis. Therefore, I researched the top 10 online newsites in Australia and based my search terms around them to see what kind of data I could draw connections from. According to Nielsen, these are the top 10 sites in Australia:

Screen Shot 2016-09-06 at 6.51.36 pm

Firstly, I used twitter advanced search to try and gage an understanding of attitudes surrounding asylum seekers and refugees within Australia. For the most part, there seemed to be around Australia’s treatment and policy surrounding offshore detention. For instance, upon the news of self-immolations on Nauru there seems to be a gathering in support of the #LetThemStay movement with many on the social media platform retweeting and “quoting” media sources who report the news with the hashtag. As well as this, when combined with words like “France” (in relation to recent terrorist actions), I started to uncover the pattern of relating attitudes to these events and certain reports. Although the reports themselves may not necessarily hold bias in these cases, it’s interesting to note the emotive reaction from twitter users relating these recent attacks to the influx of Syrian refugees – these links where often used to justify the movement of blocking off borders.

Screen Shot 2016-09-06 at 7.30.57 pmScreen Shot 2016-09-06 at 7.29.36 pm

When I attempted to narrow down searches from particular news sites, the search fields became too restrictive and I was unable to draw any conclusive data in the way that the media uses language to inform public opinion. Therefore, I begun to search for terms more loosely again, this time within social media movements such as “#bringthemhere” and “#letthemstay”. This revealed that attitudes on twitter seemed (for the most part) to be more left-wing in their thinking. As well as this, when searching these terms there seemed to be less of an appearance of news sites, whereas when you scraped twitter using the EXCEL technique with broad search terms, there seemed to be a number of people merely retweeting new reports. This showed to be that emotive movements seem to make people more likely to engage more thoroughly with the subject. From this, we can extract that this is another reason that general media outlets engage with more emotive/strong language in order to inform public opinion, as well as using emotive symbols of the movement such as Alan Kurdi to become social influencers.

Whilst I found some interesting results in scraping twitter, I do believe I would be better served to more thoroughly define my parameters when engaging with the search. Defining the tones of certain reports is near impossible in a mere 140 characters, but understanding the tone of the public in relation to certain events was very informative. However, I need to define a more accurate way to search for the way in which asylum seekers are portrayed within the media.