Leveraging blockchain power with data science
In this article Maria Weinberger discusses how blockchain offers a framework for new data monetization opportunities as well as safer, more transparent transactions.
Articles about blockchain, big data and data science have flooded tech news for some years now. The real question is what the link between these technologies is. How can these be used to create value, decrease transaction costs and increase transparency and reliability? Also, are there other potential applications of looking at data from existing blockchains?
However, first of all, let’s understand the working model of blockchain. References to bitcoin will only be used as examples, to avoid the confusion that blockchain technology only applies to cryptocurrencies and specifically bitcoin.
In fact, blockchain is an alliance between a network of computers, called nodes to collaborate on keeping a shared ledger. The aim is to keep an updated version including all transactions that have been verified by the majority of the participants. This verification step is the most resource intensive since once approved; an operation is impossible to delete or modify. This is why blockchain is considered extremely safe and reliable.
How Can Blockchain Data Become Valuable?
Since it is a ledger, blockchain contains vast amounts of transactional data which can be analyzed to extract patterns, as a specialist from data science company InData Labs explains. This could be helpful for predicting sales trends, preventing fraud or even studying social events.
Since once a block has been added to the ledger it is impossible to change and has passed a thorough verification process, it offers a seal of quality and reliability. Through its public nature, blockchain provides anyone the opportunity to trace any specific transaction or all the activity of a certain participant. This could be used to create scores for merchants and to have on public display the reputation of vendors.
In a time when data is becoming a form of currency, any blockchain can be considered a vault. Estimations go as high as $100 billion in annual revenue by 2030. The way to get the info out of the rough recordings is to motivate people to perform data mining. There is even a crypto coin, called SpreadCoin which aims to reinvent bitcoin mining into data mining. They seek to create a “Big Data Market”: a layer built on bitcoin mining.
Machine learning algorithms need data to learn from. An adequately labeled blockchain record can act as training material for such algorithms. This approach speeds up development and prevents teams from reinventing the wheel or making the same mistake twice. If a group of researchers has already used a set of data to solve a problem, their findings can act as the base for further developments. To make this into a workable model, financial incentives are required. Groups that have data can sell it on a platform and receive proper compensation.
The beauty of blockchain is that due to its open nature it can offer any contributing party the ability to verify transactions by studying patterns. When a participant detects abnormal behavior of the ledger, all other parties are notified immediately, and the malicious node is removed. The highly distributed nature also means that no one actor has the potential to make a significant impact or cause a real problem. There is only one situation when it could become an issue, and that is the case of mining pools which can have considerable computing power.
Predict Social Data
Our transactions reflect our behaviors. Therefore, by looking at data from the ledger, one can identify social trends and power centers. As generations change, the users of digital instruments like cryptocurrency, smart contracts and other artifacts built on top of blockchain will be able to accurately predict social patterns. The logic behind this relies on the fact that cryptocurrency users are roughly the same as social media users, trades are mostly by individuals, not organizations and the value is only influenced by demand.
Potential Blockchain Drawbacks
Like most emerging technologies, blockchain has a few drawbacks that need to be considered before investing.
The technical limitations derive from the cost of the electrical energy necessary to power blockchain. With the ever-increasing complexity of the problems to be solved, the amount of electricity required can be a real concern. For example, the bitcoin blockchain is expected to need more power than Argentina, which makes it unsustainable in the long run. Another technical limitation is the latency of the verification process.
A second problem has to do with the acceptance and adoption of this new approach. It has yet to make the transition from the stage of innovation towards mainstream acceptance. Most companies are aware of its existence, but few can see immediate applications of blockchain for their daily operations or are willing to invest. For now, organizations are waiting for pioneers to make the first mistakes and to learn at the expense of others.
The novelty of blockchain means that there are pending security and privacy issues to be solved. Although the underlying algorithm is deemed one of the safest possible, the connecting APIs and data processing can be subject to hacker’s attacks. A connected problem is the lack of regulations and compliance rules in this area which makes it challenging to use blockchain models for more conservative areas such as finance and healthcare.
New Business Models
The most significant advantage of blockchain is that it creates a framework to develop new data monetization opportunities. Access to transaction-level data could mean that companies will pinpoint their marketing efforts, reduce costs and increase revenue.
Looking at the ledger and tracing transactions back to their origin could mean better supply chain management, enhanced trust and even identifying vendors who underdeliver. For example, adding a “return” flag to a transaction could help build scoring models.
Blockchain is also the best environment to create data markets. Here, companies could leverage their data in any way they seem fit and create additional revenue streams.