Something Like is a community-driven, wiki-style platform for collecting and searching for recommendations on anything. It aims to replace typical "personalized recommendation systems", where driving conversions (i.e., selling stuff) is king.
Hey there, ProductHunt people!
I'm Vasily, the creator of Something Like.
Imagine this: you've just finished an amazing show and want something similar to keep the vibe going. You head to [well-known movie index], but the recommendations there somehow don't quite hit the mark...
Or this: you're looking to buy something new and seem to find an option that fits, but you want to explore other similar choices. You search for them, but find page after page of clearly paid-for "comparison" lists...
Or even this: you’ve created something, that you would like to share with the world. You know what it's similar to, you know the audience of those similar things would be a great match for your thing too, but don’t have a big marketing budget...
If any (or even all) of the above sound as relatable for you, as they are for me, you might want to check out Something Like. It's a community-driven, wiki-style platform designed to collect, refine, and search recommendations for absolutely anything! Users can easily add items, vote on similarities, and create lists—whether for personal use or to share with others. And it is the similarity, that is driving the results, not a super-secret definitely-not-sales-driven algorithm behind the scenes.
It is just the beginning of the road, with more features, more nuances for different kind of recommendations and, of course, more things of all categories and kinds ahead! And I am really excited to be here and be able to share it with all of you!
Check it out here: https://somethinglike.it
Cheers!
@kyrylosilin Hey Kyrylo!
I don't think there's a surefire way to combat spam entirely - after all, "whatever one person built, the other can break" holds true. But there nevertheless is a set of systems in place to make it very difficult to pull off.
I have taken a lot of inspiration from Wikipedia in general, and then added some more things of my own to the batch.
Firstly, the accounts are made relatively "expensive" to acquire. There's a two step verification process, with the first step being "cheap" - email verification aka "account confirmation" stage.
After this first stage, some of the functionality of the system (specifically, creating and editing things) is still locked away, but the rest is open. In this state, the account's interactions are judged differently by the system, and are there to see, whether it is a "real" user, or it's a freshly created bot/spam account. Basically, all actions of such a user are trivial to revoke, and a combination of automatic system and sometimes manual intervention can be a trigger for that.
The second stage aka "account approval" is when the user can start to do anything on the platform and could cause harm and spam - but even at this stage, it is rather trivial to revert all of the changes, because of the internal structure of collected data. If the system detects, that account is misbehaving, an alert is raised and this account could be "rejected" - basically, banned with reversion of most of the changes done.
On top of the above, for editing things, the edits are applied immediately and are show immediately, but the edited fields and values are marked for everyone to see, compare values and vote. If the edit is rejected by the community, one cannot repeat the same edit for some time - which also increases with each rejection.
And then there's this problem of bias. The system underneath is designed in a way to find a "community average" of biases of all people involved. Every user can and should freely express their opinion on the similarity - however, if it radically differs from the common consensus, the weight of the vote of this user will gradually decline for the category of things. Conversely, if the opinion does intersect with consesus, the vote weight increases. The idea is, that such an approach would produce experts/moderators for categories organically, over time, with little to no formal moderation necessary.
There's more details to all that, but I hope it answers your question, at least high level :)
Something Like
suco
Something Like
suco
Telebugs
Something Like