[Proposal] DFR - Decentralized File Rating by @Oceancap

OceanDAO Round 1
#1

Key Project Data

  • Name of project: DFR – Decentralized File Rating by @Oceancap
  • Proposal Wallet Address (*mandatory): 0x31A7f70b8812daEf0A6C7F19575DB09f726F0DD1
  • The proposal in one sentence: Oceancap is aiming towards a decentralized file rating mechanism on the official Ocean Market.
  • Which category best describes your project? Pick one or more.

    • Build / improve applications or integrations to Ocean

    • Build / improve core Ocean software

Project Overview

  • Description of the project: With our DFR solution we are aiming towards a mechanism where users are able to give a file rating after they have bought it. Later on we are integrating the DFR into our Oceancap solution. Users will be able to compare all the existing datapools additionally by the file rating score.

Website(1)
Website(1)1920×1018 259 KB

Decentralized File Rating integrated into the Ocean Market

CapCapp
CapCapp1368×769 152 KB

Oceancap App with integrated data file rating

What problem is your project solving? So far users are not able to estimate the quality of a dataset on the Ocean Market. With DFR, users are able to give a data file rating after they have bought it. Other users will than be able to see the data file rating. This will bring a lof ot value to the Ocean Market since users are encouraged to buy good datasets. Bad actors and malicious datasets will also be easier to track.

  • What is the final product (e.g. App, URL, Medium, etc)? Our final product is an integrated tool into the official Ocean Market.

  • How does this project drive value to the Ocean ecosystem?

-ensure data quality

-build trust for data pools

-attract more Ocean Market Users

-more transaction volume

-less scam

-more transparency

Project Deliverables - Category

IF: Build / improve applications or integration to Ocean, then:

IF: Build / improve Ocean core software, then:

  • A PR will be made to these Ocean components: Ocean Market Core Software
  • I commit to working with Ocean core developers to merging the PR

Project Deliverables - Roadmap

What is the project roadmap?

Phase A: 12/20–03/21

-finish DFR concept -add data file rating category’s

-work closely with community to implement new ideas

-implement “files sold” into the Ocean Core Software

-integrate DFR into the official Ocean Market

Phase B: 03/21–06/21

-integrate the DFR rating score into Oceancap

-Oceancap app development

Project Details

If the project includes community engagement:

  • Running the campaign on social media to collect ideas for rating category’s
  • Collect feedback for DFR mechanisms and Oceancap integrations

Team members

For each team member, give their name, role and background such as the following.

Benjamin

  • Role: Oceancap CEO
  • Background: PhD Business Management, Blockchain Entrepreneur

Matthias

  • Role: Oceancap CTO
  • Background: Developer
1 Like
#2

Hi @Oceancap. OceanDAO is excited to have your proposal! Might you have links to any team member profiles? Twitter, LinkedIn, Github, or otherwise? Could be useful to update in the proposal. Thanks!

#3

Right now we decided to stay anonymous, but we might change that in the future. Thanks for your understanding!

#4

For low-cost datasets, pump-and-dumpers could just invest when the price is low, buy several copies, give good ratings, raising the price, then they take their profit. The dataset might be total garbage.

The mere fact that someone bought a dataset is not enough to establish the credibility of their rating.

For a high-cost dataset, few people are going to buy it, so there will be few ratings. Rating a dataset based on the description and data sample is like rating a movie based on the poster and the trailer.

Moreover, if someone bought a good dataset for a high price and got a lot of value out of it, what are their incentives to give it a good rating? If they don’t want their competition to buy the dataset, they might say that it’s crap.

A better solution would be for there to be independent and neutral ratings specialists. Such a specialist could contact a dataset provider and vet them. Maybe they meet with them and come to a thorough understanding of how the dataset was produced. Maybe they spend some time at a computer with someone else running it, exploring the dataset, to get a sense of what’s really in it. All of this takes time and money to do, so what is the incentive for the ratings specialist? Maybe their ratings summaries are free but if you want their detailed report on a dataset, you have to buy it?

Part of the credibility of a rating comes from the credibility of whoever is doing the rating. In other words, if you were to construct an overall score, it should give more weight to ratings from people with high credibility (and low weights to anonymous ratings). In other words, there is a meta problem here: one must also rate the raters.

Look into the ratings systems at Amazon, Goodreads, Yelp, The Wirecutter, Consumer Reports, and Moody’s. How credible are they? Which ones are more credible and why?

#5

For low-cost datasets, pump-and-dumpers could just invest when the price is low, buy several copies, give good ratings, raising the price, then they take their profit. The dataset might be total garbage.

-since the datafile is cheap, a lot of users might buy it and rate it, this will counteract such behabiour

For a high-cost dataset, few people are going to buy it, so there will be few ratings. Rating a dataset based on the description and data sample is like rating a movie based on the poster and the trailer.

Users rate it after they have bought the file, so they can test more than just the sample!

So far this is an early concept, we will collect more ideas and might add rating categorys if the community wants to do so! For us a rating mechanism seems very helpful because right now there is no way to visualize data quality/a file review. When the Ocean Market is growing there might be a lot of users who are not part of the telegram community. They need mechanisms to ensure and track the quality of data files. Thanks for your feedback sir!

#6

If a dataset is cheap, that doesn’t mean a lot of people will buy it. For example, I could sell moldy bread on Amazon for one cent per slice, but that doesn’t mean a lot of people will buy it.

Datasets aren’t like books, movies or music. They’re not pop culture items with high potential demand. They’re niche products. For a typical dataset, I think it might be quite common for pump-and-dumper demand to exceed legitimate user demand.

And remember, even a legitimate user might give a good dataset a low rating because they want their competitors to think it’s crap.

I’m not against the idea of ratings. In fact, some system of rating and reviewing datasets is probably necessary, but maybe it doesn’t need to be built in to Ocean Protocol or Ocean Market. That would make it even more decentralized.

#7

Actually, isn’t staking a dataset supposed to be like rating a dataset (so that a higher stake signals higher quality, or at least a bet on higher quality), and isn’t staking already built in to Ocean Market?

I still don’t really understand how one gets enough information to decide which datasets to stake. If the datasets are expensive, one can’t buy many to look inside them. And even if one does buy some datasets to look inside them, how does one assess their potential value without having deep knowledge of the data’s potential uses? For many datasets, that’s esoteric knowledge.

#8

Actually, isn’t staking a dataset supposed to be like rating a dataset (so that a higher stake signals higher quality, or at least a bet on higher quality), and isn’t staking already built in to Ocean Market?

There are “Buy only” pools without any liquidity. How you want to check data quality for those?

Liquidity pools also could get manipulated by whales with bigger Ocean amounts. So a rating system might be needed to check the data rating score and build confidence for new users.

#9

Hi, A community member (myself), who contributes to Ocean-related initiatives, won your logo competition but you failed to provide with the prize (which was a token amount in itself). Why do you think the community should reward and trust you with 10,000 Ocean Tokens!?!

Also, can you stop using the logo.

#10

Dear Paul,

you have send us your ERC20 adress just recently (2 days ago).
and have already received your reward by now.

Check: https://etherscan.io/tx/0xb16c09326c22b3ad39956cf4f986be84dc2624dd6e59ba2f3da108c9791dcc85

Thank you!