After spending some time on LinkedIn, I felt that a map was missing to have a clear and straight forward visualizations of the Swiss employment market. This repo contains the data (scrapped) and the code to build an interactive web app. The page contains a map and four charts to explore the Helvetic job offering during December 2020.
The web app can be found here. Please visit it and interact ! (You might experience a pretty large loading time : the server as the quit sleeping mode).
First, LinkedIn was scrapped every 3h (thanks to a Raspberry Pi) to make sure that none job ad was missed. Job description and meta data was then collected, processed and stored in a more effective format (panda.DataFrame). The key information are : title, job function (or category), location and publishing date.
The ultimate goal is to have a geographical overview of the Swiss job distribution over cities, cantons and functions. Hence, the first element is a map with highlighted cantons’ borders and accessible city coordinates. Then, a pie chart would represent the proportion of job functions and two bar charts would depict the most area (city and canton) that offer the most jobs (overall and per function).
The web app is implemented in Python with the Dash library. Coupled with Plotly plots, the app is highly interactive offering a statistical description of the Swiss employment market under different perspective.
The project was design to answer questions like :
- What part of Switzerland hires the most ?
- Where could I expect to find companies specialised in biotechnology ?
- What kind of jobs are proposed in the canton of Bern ?
- Does Geneva offers non-profit organizations jobs ?
In few clicks, you might be able to get a description of where certain jobs are more likely to be proposed.
PS: futur extensions : access the jobs that are stil/currently proposed for each city, develop a temporal analysis, analyse the job description (descriptive NLP)