The Data Daily

How Blockchain is Making Data Predictions More Accessible

Last updated: 04-20-2019

Read original article here

How Blockchain is Making Data Predictions More Accessible

Thanks to developments in big data, artificial intelligence (AI), and machine learning (ML), predictive analytics is starting to become highly reliable. It’s easy to notice how Google’s search suggestions or Amazon’s recommendations seem to be reading users’ minds. Such level of accuracy is made possible by developments in predictive technologies.

These are reaching more organisations too. Prediction tools that used to be exclusive to tech companies and research labs are now being offered as-a-service. Enterprise solutions providers like IBM and Microsoft have business intelligence offerings that include prediction features. IBM’s Cognos and Watson Analytics, for example, offers both big data and AI functionalities that enable enterprise users to generate better predictions.

The problem still is that these tools are mainly accessible to large enterprises who can invest in tools, time, and human resources. Effective predictions require well-trained data scientists who would work and iterate on models. These could also take weeks to accomplish. Given the resources needed, smaller business and organisations are often unable to benefit from the technology. 

Fortunately, blockchain, as one of today’s key emerging technologies, offers the ability to democratize access to such capabilities. Ventures like Endor, Augur and Gnosis, seek to leverage blockchain’s strengths to create prediction platforms and ecosystems that make accurate forecasts accessible to a wider set of users. Here are three ways blockchain is making this possible.

There must be a conscious attempt to democratise predictive technologies. But for any such effort to succeed, it must be able to lower the barriers to the required expertise and resources. Predictive analytics typically need massive volumes of data, data science and AI expertise, and computing power to perform.

Endor seeks to address these concerns. The company isn’t exactly a new player in predictive analytics. It is an MIT spinoff and core of its technology is Social Physics. Social Physics is a body of knowledge that promotes the use of big data to understand human behaviour. It allows for the detection of emerging behavioural patterns from human-events datasets. Using Social Physics, Endor has created what they refer to as the “Google” for predictive analytics. Users can simply ask predictive questions and readily receive accurate predictions in return.

Endor has already made this technology available for large enterprises, but the team wants to provide the same benefits to everyone. To democratise access to AI-driven behavioural prediction, the company has turned to blockchain and is set to launch its Endor.coin protocol. The decentralised and trustless protocol facilitates an ecosystem of data and engine providers, developers, and users who need predictions. 

Essentially, the protocol works like a platform-as-a-service, but instead of being controlled by a centralised authority, it relies on an ecosystem of participants. It also uses its own EDR token to fuel transactions. Users and data providers pay using EDR to ask predictions and have their data analysed respectively. Developers can earn EDR by writing reusable software components. Through this ecosystem, the participation of more stakeholders brings down costs and leads to more efficient engines.

Another way blockchain technology is making predictions accessible is by powering prediction markets. Prediction markets are platforms where parties could wager on the outcomes of future events. Common use cases for prediction markets include price discovery and financial markets. 

While Endor relies on Social Physics and AI, blockchain-driven prediction markets essentially crowdsource and rely on the wisdom of crowds to reach predictions. Currently, there are two such prediction market platforms worth mentioning – Augur and Gnosis. Both offer similar functionalities including supporting binary, categorical, and scalar outcomes. Both platforms also use crypto tokens as means to facilitate transactions. 

These platforms are also built to be decentralised and secure bringing trust in the system. Blockchain’s transparency discourages manipulation. Anyone will be able to review the transactions. Agreements made using smart contracts guarantee fulfilment. These mechanisms guarantee a fair playing field which encourages participants to apply the best prediction models to secure their wins. Depending on the nature of the wager, others can also then use these predictions as a reference to guide their decision making for various other purposes.

One of the key challenges in prediction is getting hold of big data to fuel analytics. However, blockchains themselves are sources of data. Public blockchain protocols like Bitcoin and Ethereum have already recorded transactions in the hundreds of millions each.

Much of the information stored in blockchains may be encrypted but there are also plenty of other insights that can be drawn from the available information. This limitation isn’t stopping efforts from putting blockchain data to good use. Tools can also Endor, for instance, supports the analysis of blockchain data. 

The rise of blockchain platforms like Ethereum has even upped its potentials as a source of rich insights. It is being used by various decentralised applications (dapps) to record transactions. These dapps already cater to a host of industries including finance, healthcare, and marketing and the patterns from these transactions could be quite valuable to their respective verticals. In addition, blockchain’s immutability guarantees the integrity of the information. The availability of varied and quality data can allow more accurate prediction models to be created.

These various blockchain efforts are creating experiences that enable more people to have easier access to predictions. The ecosystems these services nurture also encourage participation. The sharing of more data and the development of better predication models benefits all stakeholders. Ultimately, these predictions and insights are to be used to make effective decisions which could only lead to better outcomes for everyone concerned.

Read the rest of this article here