Model Description LDA is a topic modeling technique that uses a set of assumptions about how the writer generates the document. First, it assumes that the person is going to decide on the number of words according to a descrete distribution.
Second, a topic mixture is chose based on another distribution. Each word in the document is generated by first picking the topic and then selecting the word. LDA back tracks this process.
LDA creates a priror assumption and randomly assigns words to a topic. Then it improves on that based on a mathematical equation to change the topic of each of the words one by one. Eventually there
Solutions Expand use of State programs to incentive employment of former offenders Implementing the Fair Choice Ordinance which has had a successful track record in San Francisco of promoting employment of former convicts Expand health and human services in the areas that need it most. Currently health and human services offices (mental health, rehabilitation, goverment housing, etc.) is not located in the areas where it is needed most Continuous monitoring of the voice of the people in need through Street Wise topic modeling