Scraping Big Data
This post became more relevant than ever after the Cambridge Analytica Scandal that broke on March 2018
We may not know who is going to be the next president of the United States, but we do know who’s going to be the real winner of the race: Big Data.
Every day candidates from both parties are monitoring millions of potential voters’ profiles with automated social scraping tools, in order to learn more about potential voters and improve their campaign strategy.
It is a new, game-changing frontier. Yet, most voters and political analysts are not yet aware of it.
Most Americans are using social media to broadcast their thoughts and views on a daily basis. Social networks have therefore quickly turned into huge databases, where one may find useful demographic and psychological information using the right tools.
By monitoring online profiles through data scraping companies, US campaign strategists can analyze the interests and behaviors of millions of floating voters. They are learning about their nominees’ momentum and weaknesses from the reactions to statements and actions, even before they get covered. Big Data is a new, smart method to avoid future mistakes, enact damage control, and boost successful strategies.
Along with the masses, social media monitoring companies are keeping an eye open on influencers, those who boast very large numbers of followers, readers, and fans on social networks. Just a few, strategically placed actions, in fact, can affect the course of a campaign! Candidates are now using social media monitoring to decide on which issues they will cover and which narratives they will use in their future public appearances.
Technically speaking, how can the social data aggregation companies collect Big Data?
They can either get access to data provided by the social networks through their API (i.e. get the official data through a programmatic interface). But the official data are often incomplete, so many social media monitoring firms are enriching their data by harvesting information from blogs and the social networks themselves to get a more complete picture. These sites are not eager to allow access to their information – even though it’s arguably created and owned by the public; access is often blocked and data may be falsified to counter automated data extraction.
To avoid it, these companies use proxy networks, which rely on IP addresses located in huge data centers. The proxy solution allows data aggregators to scrape information with more ease, but not necessarily with confidence about the results. In fact, IPs in data centers are easily identified as proxies.
That’s why top tier firms use peer-to-peer proxy networks, such as Bright Data, to route their requests through residential IPs. These IPs cannot be identified and make the process possible.
However, when harvesting for data, the companies should always make sure the information is publicly available and legal to access. Otherwise, as happened with Ted Cruz, the aggregation is likely to backfire.
The insights being gathered by campaign strategies on the floating voters are larger than ever, and unlike in the last elections, they are going to play a major role in the decision-making process.