Our "Hive Discover" service has been available to everyone since last week and the first users have already had content recommended by our AI. Unfortunately, we do not know how their user experience was and how they think about the suggested content. That is why we are calling for a big feedback round: We are looking for interested people who have some free time and enjoy testing new things. We would like to ask you to use our service and think about topics such as user experience, the relevance of suggested posts and functionality. After that, a simple comment below will suffice. But if you want, you can also write a separate post and tag us. We will take the comments/posts to heart!
Link to our service: https://beta.hive-discover.tech
What is „Hive Discover“?
Hive Discover is a new service in the HIVE Ecosystem to help find or discover content. For this, we did not just develop a blunt search engine. Instead, we created an AI that does the job. This AI analyses any post on the HIVE blockchain and suggests qualitative content to you. The information for this is simply read from user activities like recent votes or posts.
What data do we use and how do we use it?
In general, we can proudly say that we only store the most necessary data. This starts with the login, where we only ask for a key to verify your identity using Hivesigner and do not ask for permissions to post or vote, for example. Furthermore, we only store data locally for the last 100 days. This means that all posts are automatically deleted after this period. We do store accounts permanently, but in two different states: On the one hand, we have stored the general account; more precisely, the username and the profile information (name, location, description...). We need this for the account search. On the other hand, we save the last votes of active users, i. e. those who have registered, because these are needed by our AI. But here, too, we can only find votes from the last 100 days. You can also say at any time that you want your data to be deleted. Just go to the settings.
Those who do not want us to have information about them can also enter this in the settings (at a later date, this setting option will be optimised with the help of Custom Json Operations).
What could the future look like?
After this feedback round, we will first look at what we should improve. However, this could take some time. After that, we plan to improve topics like the suggestion algorithm and text analysis. We also plan to provide an API that would allow Front-ends to interact directly with our service. As a result, the user experience could increase dramatically because you would no longer have to switch websites. However, it is not clear at the moment when we will get to work on this.
What is clear, however, is that our service will remain free for everyone. Even when the API comes, it will be free to use!
Stay healthy, Team Action-Chain