Repository navigation
0. About
ReMatch is a web application that matches employees and recruiters by use of the Elasticsearch data analytics engine which determines a compatibility score based on their individual profiles. The users are able to view matches that have the highest compatibility to their own criteria, see the detailed information of their matches and express interest if they choose to.
Visit our application at: rematch-htw.herokuapp.com
Privacy information: For now this is just a university project that does not adhere to data protection guidelines. The name and email you add can be revealed to anyone with an account. You can choose to enter mock data or an invalid email address to avoid this.
- On the index page choose to sign up as a job seeker or recruiter.
- Answer the questions about you and your expectations.
- Create an account with an email address and password
- Log in using your credentials
- You will immediately get matched with jobs or candidates in our database
- Click to view the details of a match and express interest through a message
- If there are no satisfying matches, you can change your preferences in your profile or come back later to see new matches
- As a job seeker you have the option to search for jobs freely
- Recruiters have the option to create more jobs
We use Elasticsearch (ES), which is a popular information retrieval system. With ES we can retrieve data with customized parameters and it will show us the best results based on our query. ES has an internal algorithm to calculate a score. The higher the score is, the more accurate the retrieved information is.
The information from your profile is used to create a query for ES which will find the best matches sorted by their ES scores. During the sign up you will also receive a score based on your skills and expectations. This is what we call a maximum score. It is calculated by creating a fake perfect match in ES.
The ES scores are relative to each user's profile and therefore difficult to assess. This is why we display them in the frontend as words like "Great" and "Good". To determine what makes a score great we compare the individual score of the match with the maximum score.