The data-conscious organization’s solution to fake news: moderation, management and transparency
“In god we trust, all others must bring data,” a line attributed to the American statistician, W. Edwards Deming, highlights the importance of statistical measurement and analysis in verifying facts and confirming their viability. However, as the volume of data grows exponentially, this becomes far more difficult with each passing month.
While the fake news epidemic may be most noticeable on social media platforms, businesses face a similar challenge internally. Misused, fictitious or ambiguous data can lead to serious errors in decision making, as well as reputation damage, and breaches of legislation that lead to costly fines. Since the General Data Protection Regulation (GDPR) came into force , this has become increasingly important, as the legislation increases the onus on companies to make sure they know what data they store, where they store it and whether they are using it in a compliant, secure fashion.
Unlike fake news intended to sway public opinion or for comedic purposes, not all misinformation is driven by malicious intent. In a business context, for example, it could be that those responsible for developing the insights have insufficient data, insufficient skills, or because they are intent on proving a pre-existing hypothesis.
The good news is that these oversights can be remedied with a three-pronged approach: moderation, management and transparency. This is where the emerging roles of the data curator and data protection officer (DPO) come in. Just as a successful museum curator sorts through masses of exhibits and information to pull together the right pieces to tell a story, so does his data counterpart, who curates data to be a successful driver of business decisions. The DPO ensures governance and promotes responsibility. In the era of GDPR and data-driven business, the time has come for businesses to own responsibility for the data they hold, and how they use it.
Under GDPR, companies have to be transparent with their customers about what data they have about them, why they have it, and what they are doing with it. Ensuring that the file metadata is kept up to date with the correct data lineage will be crucial for organisations. This is because it will allow organisations to catalogue any revisions, noting when and how the data was created and changed. This will be invaluable for reinforcing the integrity of data analysis by providing a clear modification history – allowing any unexpected changes to be monitored and amended.
This ethos can be seen in Wikipedia’s operation model, which is based on open collaboration with transparent records of what content has been changed, by whom and when. The result is a carefully decentralised service which seeks to consistently provide reliable and balanced information. This is an ethos that businesses can learn from in their quest to remove bias and siloing from within organisations – creating a shared neutral platform of carefully curated data.
Managing data responsibly
However, even with fully transparent data management, oversight is still needed to make sure that the changes being made fit within the structures and rules of the organisation and industry. This is where a “self-service governance” model is important – allowing users to add content, consume it and edit it, but within the confines of an established set of rules that the data curator can oversee. In other words, everyone is responsible for ensuring data accuracy at the employee level. This ethos combined with the data curator’s framework of rules ensures that data is kept accurate, labelled and tested, which is key for data discovery as well as analysis and data enrichment.
Similarly, with the vast amount of information available to today’s businesses, the data potential can feel like a data overload. With the rising decentralisation of information within organisations thanks to the proliferation of resources like the Internet, employees often need to pull from multiple channels of information which need to be managed and collated effectively before they can be used. The data curator is an essential part of this process.
The GDPR’s requirement for many larger companies to appoint a data protection officer cements this into the fabric of organisations – creating a central figure who informs and advises the organisation on data protection compliance. This central figure who is tasked with ensuring the ongoing validity of data utilisation is a valuable asset for companies of all sizes looking to make reliable, actionable decisions based on data. Well, that’s the theory at least. According to the Talend’s second annual GDPR Benchmark research, over half of companies surveyed were not able to meet data access and portability requests within the GDPR-specified one-month time limit. In practice, companies do not adequately track personal information and often nobody has been mandated in a tightly-defined way to manage the processes, to set best practices and provide the technological solutions.
Having a DPO in place can also transform GDPR into a business driver and a competitive differentiator. The DPO will put in place the necessary strategy and processes, work with the IT department to select the right tools and will bring all the data workers and business users together. As a result, he or she will ensure the company has a strong sense of data ownership. This ownership leads towards a common sense of data protection and a general agreement on the data governance strategy to put in place. The appointment of a DPO is the first and probably the most important step towards a well-governed and responsible data management strategy.
Moderating the data process
Overseeing the data process is one of the core responsibilities for data curators. In the social media world, we see companies like Facebook employing over 4,500 content moderators whose job it is to monitor user-uploaded content for any unsuitable material. In the business world, this will mean ensuring the data that is uploaded, amended and shared is kept up to date and accurate.
One of the areas where moderation will be particularly important is with the increasing use of AI and big data analytics within organisations. Inputting historically inaccurate data into the program will cause inaccurate suggestions for the future – in human resources, this could manifest as bias, in sales, this could mean one product being mistakenly targeted at the wrong audience or to the detriment of other offerings. A human operator who is responsible for spotting anomalous results, and checking the data that informed these findings, is crucial for ensuring that decisions based on data are reliable and accurate. This ensures that data offers businesses a competitive edge and business advantage.
The impact of successful data curatorship is hard to overstate – surveys from the likes of CrowdFlower estimate that 80% of data scientists’ time is spent on preparing data for analysis, compared to just 20% on analysis. However, with a data curator on board, organisations could significantly reduce this time by improving data quality, data accessibility and data integrity. As a result, organisations stand to benefit with class leading, industry compliant insights, rather than being overwhelmed by the data tidal wave.
So, becoming more data-conscious and responsible could be thought of as adding a Data Curator and Data Protection Officer. If we assume that the DPO is also responsible for driving data awareness, the problem is partially solved. Partially, because it is a collective effort which can be impactful if - and only if - people have sufficient knowledge about the data issues to question them.
This article was originally published in Data Economy on December 12th, 2019 https://data-economy.com/the-data-conscious-organisation-and-fake-news-solved-by-moderation-management-and-transparency/