Topic: [ANN][preICO] Dolphin BI: The first cryptoasset investment analysis marketplace - page 11. (Read 5235 times)

Yes, openness is a core principle of our platform.

Anybody can become an Expert but the rating system will help to choose best of them by the community voting.

Also anybody can become an Author (thanks to widget container's isolation) and will get access to basic resoures of the platform. With the rise of popularity of Author's widget it will get more computing resources automatically.

Can anyone become an Expert or an Author on your platfrom? Or do they have to be tested and will you ask them for qualification proof?

Thanks a lot for warm words! 
We would be very greateful if you could recommend our project to someone else who could also be interested in.
We also hope to see you as early backer and contributor! 

this concept is very interesting and very smart idea, I am very interested to join this project, I will continue to see the progress of this thread, hopefully run very well

The authors are people responsible for widget creation. They are usually developers (they have to create a widget using an SDK, connect it to a data source and register it on the platform. After they register a widget, they get a set percentage of the widget's revenue (or all of it, if it's automatic).

The experts, on the other hand, provide actual content, if the widget is not automatic. They can write business plan overviews, technical audit reports, etc. in appropriate widgets. The part of the widget's revenue not going to authors is split among them according to their reputation.

With independent Experts we target the following:
1) make Experts independent from the platform and allow anybody in the world to monetize their analytics skills in part of ICO analysis
2) create a modest competition between Experts

We expect to fight controversial situations by fine tuning rating system.

In your opinion, wouldn't the Expert's financial reward dependency on Expert's grade cause many controversial situations? I am also curious why you preferred this way of analysts monetizing their expertise instead of e.g. just hiring couple or team of experts.    

reserved the Chinese translation

Thanks for a good question!

Our team members have a background in OLAP technology and in Machine Learning to bring automatable analytics whe possible.
Also we implement Experts' evaluation method to assess characteristics as market size, team skills and experience, competitors which cannot be evaluated by machine right now. We already have preliminary agreements with some ICO experts to become the first users of our platform.

It's our inportant statement: we as a platfrom developers don't evaluate ICOs by ourselfs so we avoid a conflict of interests. It differenciates our service from any other rating agencies and ICO listings.

We use a token as an utility token:
1) users pay for the service with token
2) Experts and Authors receive a reward for ICO analysis contribution with this token
3) token spendings and daily average balance is used to prevent rating system attacks

So we sure a token is a crucial part of our system.

Yes, a ticker and a smart contract address will become visible for all on presale site when public TGE will start (on Monday, 18 September 16:00 UTC).

Thank you guys for the clarifications! Another thing in your project I am curious about is the authors and experts. What are the benefits for them and what would be the ways to attract them? As I understand from the whitepaper, they have to create widgets, does that mean also the technical part (programming), or authors and experts provide analytics and content only?

reserved.I am more interesting about the bounty thread

Hey, where can we publicly track the amount of presale contribution? Is it even public?

What skill set or backgorund do the team possess to provide a very good analytical intelligent investment advise for me. I don't think their is need to have a token for this because is more of service subscription for me

Of course the bug-bounty program will be available for each part of our service. Because the smart-contract core is what operates with tokens it should be protected with highest priority. Other part of the platfrom as frontend or data providers could also contain security flaws so bug-bounty program will be definitely available for them too.

It will be soon, if you maintain the Dolphin project))

waiting for bounty time. 

Hello Elementique!
Actually, the project itself started from analyzing distribution curves of ICO payments. We have found that for all solid ICOs the distributions after logarithmic transformation looked very similar (bell-shaped, so the underlying distribution could be considered Gaussian or Student's). This allowed us to detect tampering in payments - for example, the project investing big in the start of its own ICO to simulate high demand. However, the main problem here is that this analysis can be only performed after the ICO, so we decided to instead concentrate on something more useful at all stages, like sentiments or ICOface (a database of all people who did an ICO, where you can look them up by their photo).

So what you can surely expect in the future is payment curve and payment distribution analysis, and we also had a cool idea floating around about clustering addresses to find out whether, for example, exchanges were involved. But this is not our focus right now.

As for detecting whether the ICO is fraudulent or not directly, several problems arise - first, the data on fraudulent projects is scarce, not nearly enough to train a decent model; second, people seem to distrust even an amazing model, when the task is formulated as predicting fraud/not fraud. And if you make even a single mistake, you can really upset either investors or the project team that you called fraudsters. After interviewing some investors we found out that they would prefer to see indicators of possible fraudulency themselves and pass judgements on their own.

In the meantime, we will concentrate on some basic widgets like exchange rate curves, number of token holders, token balances distribution and so on.


This is going to be an automated bug hunt after the initial version of Dolphin BI smart-contract will be deployed. The idea is that we set the contract's invariants, and people who manage to break them will automatically be rewarded.
Here's an article where you can learn more: https://blog.zeppelin.solutions/setting-up-a-bug-bounty-smart-contract-with-openzeppelin-a0e56434ad0e

And perhaps could you please give more insights into the "bug-bounty" program? How it will be implemented, and what would be the main purpose of it? 

Thank you for your answer! I have couple of other questions: you have already mentioned some resources that you are collecting the data from, like social networks. Could you please name some other sources of the data you are using or want to use for analytics? In your opinion, what would be the most crucial criteria (one or several) in order to identify fraud/suspicious project, and would it be possible to do it based on automated analytics only, without expert involved? Speaking about investment analytics, in which form the final automated investment recommendations/fraud warnings will be given? Will it be some sort of ranking?