Discussion 5
Discussion 5
in which they exist. Using a set of images, Bhadouria et al. (2017) detailed the process of using
an algorithm to classify businesses advertising jobs into the relevant industries. Advanced UML
A typical job posting is shown in Figure 4, which also details the process that will be utilized to
carry out feature extraction. For example, a code breakdown could be incredibly helpful for
developers, but it has been excluded from both the training and test models. In Figure 5, you can
see the whole machine learning model, including all of the steps, and how to use it to classify job
postings. LinkedIn's approach to diagram creation departs from the conventional UML structural
diagram by omitting the coding breakdown that would be useful to developers. The utilized
graphic displays more of the data that the machine learning would produce, rather than the
required code.
By looking at the structure of the graphic, it appears that ML is extracting existing object codes
for data instead of writing its own. Developers can save time by not having to draw as many
diagrams with new codes nor incorporate into the system being built. They use predetermined
codes to route items. There can be a lot of reasons why the patent doesn't specify or contain a
sequence diagram. This statement made it clear what the problem was and why each step was
Reference
Bhadouria, Sameer Mahendra, Aaron Tyler Rucker, and Jason Willard Carver. “Machine learning
algorithm for classifying companies into industries.” (Links to an external site.) U.S. Patent