You meet someone for the first time, shake their hand, look into their eyes and wonder:
Is this someone I can trust? Meshwith provides honest insider info so you know who to trust - who you will will meshwith at work and in your social life. No surveys or (potentially fake!) letters of recommendations, just instant honest opinions about anyone.
Where do these trust networks come from? You - the user!
Networks that store trust/distrust relationship are called trust networks, and they're different from social network because you can trust OR distrust anyone with an email address. Meshwith uses these networks to predict who you can trust based on who you current trust/distrust.
Challenges we're facing
There are some common problems with getting honest opinions from users about people they know. They're illustrated below. We have some solutions to these problems using graph theory tools and machine leanring to hide the identify of people contributing ratings, but still present you with accurate personalized data on who you can trust. Further details on the algorithms closer to launch.
What are we selling?
We're selling meshwith queries for 70% off. Each query tells you if you are likely to meshwith a single person.
- If you're just looking for a new job or a babysitter, you'll probably use a couple hundred queries.
- If you're a professional recruiter you'll probably screen hundreds of thousands of applicants with this new method, against all the employees in your company, or against the team they might work with.
Your contribution covers the cost of building the webservices, graph database, cloud scaling and management, GUI and implementing the machine learning algorithms that predict trust, on both the mobile and desktop versions.
Why am I working on this?
As a neuroscientist, it's clear to me that the long-term obstacles to scientific progress are NOT related to technology. The greatest long-term obstacles to good science are racism, sexism, workplace abuse and general selfishness among scientists who think they own data that was paid for by citizen taxes! If there is a feedback system for scientists, we can encourage collaboration with researchers who do great work, in the same way that yelp can lead you to great restaurants.
Because my research involves using big networks to find the source of disease, I designed the meshwith trust network to identify people you should avoid. There's a positive side to this - meshwith also identifies fantastic people!