sockit 0.0.2

Assign probabilistic Standard Occupational Classification (SOC) codes to free-text job titles

Sockit is a self-contained toolkit for assigning probabilistic Standard Occupational Classification (SOC) codes to free-text job titles. It is developed by Research Improving People's Lives (RIPL).

Installation

Requires Python 3.8 or later.

To install from PyPI using pip:

pip install sockit

To install a development version from the current directory:

pip install -e .

Running

There is a single command line script included, sockit, that processes existing CSV files containing free-text job titles in one of the columns:

sockit -h

usage: sockit [-h] [-v] [-q] [-d] -i INPUT [-o OUTPUT] [--record_id RECORD_ID] [--title TITLE]

Sockit: assign probabilistic Standard Occupational Classification (SOC) codes to free-text job titles https://github.com/ripl-org/sockit

optional arguments:
  -h, --help            show this help message and exit
  -v, --version         show program's version number and exit
  -q, --quiet           suppress all logging messages except for errors
  -d, --debug           show all logging messages, including debugging output
  -i INPUT, --input INPUT
                        input CSV file containing the record ID and title fields
  -o OUTPUT, --output OUTPUT
                        output file (default: stdout) containing a JSON record per line: {'record_id': ..., 'title': ..., 'clean_title': ..., 'socs': [{'soc': ..., 'prob': ..., 'desc': ...}, ...]}
  --record_id RECORD_ID
                        field name corresponding to the record ID [default: 1-based index]
  --title TITLE         field name corresponding to the title [default: 'title']

Alternatively, you can load the sockit package in a python script and process titles one at a time with the search() method:

import sockit
clean_title = sockit.clean(title)
result = sockit.search(clean_title)

License

Sockit is freely available for non-commercial use under the license provided in LICENSE.txt. Please contact [e-mail] to inquire about commercial use.

Contributors

• Marcelle Goggins
• Ethan Ho
• Nile Dixon
• Mark Howison
• Joe Long
• Karen Shen