Fig. 2: A schematic diagram of SSCN based skill valuation framework.
From: Market-oriented job skill valuation with cooperative composition neural network
a An example of job posting in our data, which consists of some structured contextual information (e.g., company name, timestamp of publishing, work location, and required working experience), expected range of monthly salary (i.e., lower/upper bound in RMB), and detailed job description that introduces the requirements on candidates' job skills. In particular, each skill usually has a descriptive requirement on the degree of mastery, such as Proficient in JavaScript, and Familiar with AS. b We formulate the job posting as a set of skills formed in a skill graph, and some contextual inputs. Our proposed SSCN estimates skill value and combines them into the job salary. The color gray, blue, yellow and pink indicate inputs, model structures, outputs, and loss functions, respectively. c The detailed structure of CSVN. d The detailed structure of ASDN.
