CrowdEngineering has put a patent on crowdsourcing. As you may remember, CrowdEngineering provides a platform designed to fully integrate existing processes and applications with the benefits of technical crowdsourcing. This week they put a ring on it, so to speak, by securing a patent that covers its core technology for enterprise crowdsourcing, the foundation of its flagship CrowdForce platform.
The Secret of Skills-Based Crowdsourcing
The patent for invention 8,099,311, “System and Method for Routing a Task to a User in a Workforce,” is for a skills-based crowdsourcing system that selects workers among community experts, taking into consideration their preferences and skills as well as the performance and quality they delivered in the past.
You might think that patenting an aspect of crowdsourcing goes against the open-source nature that matches tasks with qualified professionals, but it’s still a capitalistic society, so patents are necessary to protect one’s “secret sauce” as Gioacchino La Vecchia, Chairman and CEO of CrowdEngineering puts it.
According to the patent:
In workforces that include a large number of users who may perform tasks, tasks are allocated to the users or a group of the users. A task may include, for example, a problem to be solved, a question to be answered, a survey or questionnaire, a request for proposals, or the like. To allocate tasks to users in a workforce, a task including normalized task attributes is received by a routing system. The normalized task attributes correspond to skill metrics within a profile of each user. A supervised learning portion of a neural network calculates an initial value based on the skill metrics of each user and the normalized task attributes. The initial value associated with each profile is input into an unsupervised learning portion of the neural network having weights. In the unsupervised learning portion, the weights are based on measures of behavior of a user, expressed quantitatively by performance indicators within the profile of each user. For each profile, a neural network is generated to calculate a fitness metric. To distribute tasks among the users, a stochastic model is used to select a profile based in part on the fitness metric. The task is then routed to the user associated with the selected profile and feedback is collected from the individual and the selected user."
It’s clear that CrowdEngineering’s patent aims to redefine the unique connection between the internet and the job market by routing tasks and assignments in a crowdsourcing environment. Only time will tell if putting a patent on crowdsourcing will help or hinder the ability to get work done.