Plan to do some experiment on User Data and Job Description
Pre-Step:
i. experiment on common words
ii. extract useful column from user Data
Planned Steps
a. Use TF/IF to find out most important words from User Data and Job Decription
b. Then Use LSA (Latent Semantic Analysis) transfer these word into a fix sized vector
Future Step:
c. Buld Matrix M
i. row = job id
column = job seeker
d. Build Collaborative filtering model based on previous vector
sim(x, x') will be the product of the two job description Vector
final prediction of sim(x, y) = sigma s(x, x') * M(x, y)
where x = job ID to predict , y = recruiter
KunSu
commented
4 years ago
Planned Steps a. Use TF/IF to find out most important words from User Data and Job Decription b. Then Use LSA (Latent Semantic Analysis) transfer these word into a fix sized vector
Future Step: c. Buld Matrix M i. row = job id