Back to news

BLOCKBIRD detailed description

The rising tide of information

Living in the current digital era sure is a privilege, unknown to our ancestors. Being alive in this age alone presents you with more opportunity than ever in the history of mankind.

However, nothing ever comes without its drawbacks. Digitalization brought up another, previously unknown, issue in the form of information overload. We are bombarded with posters, commercials, and data every step of our daily lives.

Sure, we (The BLOCKBIRD Movement) are also “guilty” of promoting ourselves – that’s just how the wheels turn nowadays. Nonetheless, we believe this is the ‘necessary evil’ we have to accept to achieve just the opposite: tackle the information overload and ease the burden of all crypto contributors and enthusiasts.

So, how are we going to achieve that? This article is going to explain our way of solving the problems crypto contributors are very familiar with: information overload and scattered news that can be easily manipulated. With the current state of the news and information, decision making becomes time-consuming, and it limits anyone’s ability to understand the already-complicated crypto territory.

Understanding the news is like baking bread

To paraphrase: imagine you had to make your own bread. You would have to find the right wheat, crush it into flour, and find all the ingredients to make the dough. After that, you would have to bake it at the right temperature for just the perfect amount of time, and, voila – you have your bread bun! Sounds easy right?

Now imagine that you have a dozen different kinds of wheat popping up (information overload) all the time at different locations (scattered information sources), and a lot of them are dangerous (manipulated news).

After going through all the wheat, and (luckily) picking the right one, you would have to go through the whole baking process (extracting and evaluating the core message) to finally get your morning bread (apply the information to the market).
And the next day, you would have to do it all over again (you want to stay up-to-date, right?).

Wouldn’t it be nice to have someone deliver you fresh bread every morning? That’s where we come in!

The first step – The Gathering

We had to approach our issue systematically, and to do that, we first asked ourselves: what is the best way to tackle information overload? We quickly realized that we are not going to reduce the amount of information we see daily.

What we can do, however, is enable people to ignore the majority of crypto-related information, as most of it is duplicated, irrelevant, or of low-quality.  To achieve that we have to offer only the essence of the information, gathered from across the online spectrum, and present it in one place in the most understandable way possible. Doing so, we not only lessen the burden of information overload but also solve the issue of scattered news.

What follows? – Information processing

After we gather all the information, we have to process it in order to be able to evaluate it and further present it in a user-friendly way. To process the news we will be using the various methods of text analytics to convert the raw information into meaningful data, ready for analysis.

(The text analytics methods include information retrieval, lexical analysis (studying of the word frequency distributions), pattern recognition, tagging/annotation, information extraction, data-mining techniques (including the link and association analysis), visualization, and predictive analytics.)

Since we have the enriched data, we can use it as an input to Machine Learning algorithms that will automate many processes on our platform, which include:

– correlating the content of the news with the source rating (that enables us to understand what makes crypto news trustworthy and objective);

– analyze the content, sources, authors, and other relevant parameters in order to detect and exploit the interconnections between crypto news, or what we call, the “news graph”.

The “news graph” is one of the most important aspects of the platform, as it enables us to understand the features of crypto news and the way it is created and spread. The statistical data will also make it possible to define the characteristics of manipulated news, so we can expose them in an efficient way.

We will use the crypto “news graph” along with content analysis to:

– calculate the similarity index, detect duplicated crypto news, and make user recommendations;

– categorize news in both pre-defined and ad-hoc categories;

– extract relevant keywords; 

– create summaries of the content.

After enriching the information, evaluation follows

To effectively evaluate news, we will rely on three pillars: the rating, evaluation, and reward system. All three are interconnected in a way that enables us to provide the best possible evaluation of the news.

Evaluation 

We decided to approach the news evaluation in a different way. In order to get as accurate as possible, we plan to combine the community evaluation with the Machine Learning algorithms. The Machine Learning algorithms on the BLOCKBIRD platform will be trained and updated using the community’s collective intelligence as input, thus creating a hybrid intelligence system. This synergy offers the best combination for us to be quick and accurate when evaluating the news.

Rating 

As is the case with any community, its members vary greatly in terms of knowledge, expertise, capabilities, interests, experience, and views. It will be the task of the Machine Learning models to take all these differences into account and use them to shape the earliest, robust rules. Technically, this means that the models will calculate different weightings for each user based on their crypto news ratings, usage behaviors, and connections while eliminating the unnecessary information.

As users are not the only factor in the process of news evaluation, we will also use similar machine-learning approaches to appropriately rate the news sources. The rating system is supported by an algorithm that calculates the value of presented information through parameters such as the number of votes per information, user credibility rating, amount of views per information, and a number of shares per information.  

Reward 

Users are rewarded with BBIRD tokens in two ways:

Firstly, users are rewarded for every action that helps the BLOCKBIRD platform to display and identify relevant crypto news, find manipulated crypto news, discover duplicated crypto news, correctly rate credible sources and influencers, or post meaningful comments.

Secondly, users are rewarded for several other activities on the platform depending on their engagement (such as the time spent on the platform, number of votes and their impact, or the ratio between read and followed news).

What about manipulated news?

As we will develop one of the most advanced evaluation systems, we will have great capabilities of discovering false and manipulated information. Nevertheless, some information is impossible to be confirmed as trustworthy right away, as it comes from sources that are not easily verified.  For example, if a team declares a partnership, it is quite difficult to check up on their statement. However, we can mark it in a way that indicates that the partnership wasn’t confirmed by both parties. The marking mechanism can also be used to tag questionable and contradictory crypto news.

To tackle this issue, we  plan to develop an alert system that would warn our users when certain information turns out to be false, offering them the chance to act as quickly as possible when things turn around on the markets. The combination of our state-of-the-art evaluation system, crypto news marking, and crypto news alert system will, therefore, act as a safety net for our users when dealing with misguided information.

Conclusion

And there you have it! This is our way of dealing with information overload, dispersed news sources, and manipulated news, all of that on a single platform. As we highly value our community, we would love your feedback and thoughts on the current crypto news situation. To get involved in the discussion, you can join our Telegram  or Reddit section.  If you wish to further educate yourself on our project, you are invited to check out our blueprint and become a part our movement!

November 7, 2018, Maks Levstek