Skip to content

Instantly share code, notes, and snippets.

@fabianabel
Last active September 22, 2015 10:10
Show Gist options
  • Save fabianabel/18a8700705f53b000a46 to your computer and use it in GitHub Desktop.
Save fabianabel/18a8700705f53b000a46 to your computer and use it in GitHub Desktop.
Idomaar Dataset: Job Recommendations

Idomaar Dataset: Job Recommendations

Idomaar format

An entity will have 5 attributes:

  • type: posting
  • ID: ID of the job posting
  • timestamp: unix timestamp (in milliseconds) that specifies when the job posting was published
  • properties of the job posting (JSON-formatted)
  • links to entities: empty JSON object

A relationship has the following attributes:

  • types: impression, click, bookmark, rating
  • ID: auto-incremented ID that can be used to bring the relationships into a temporal order
  • timestamp: unix timestamp (in milliseconds) that specifies when the impression or interaction happened
  • context info about the impression/interaction, e.g. in which section was the impression (JSON-formatted)
  • entities:
    • user ID: ID of the user who performed the interaction or to whom the posting was shown
    • item ID: ID of the job posting on which the interaction was performed or that was shown

Postings:

posting	42	1404165599468	{"title": "Experienced Java Developer (J2EE)", "discipline": 1011, "industry": 20100, "city": "Frankfurt"}	{} 
posting	70	1404165654321	{"title": "Software Engineer (Scala)", "discipline": 1012, "industry": 20100, "city": null}	{} 

Relationships - Version A (plain array):

impression	1	1404166899123	{"section": "jobs", "source": "recommendations"}	{"postings": [42, 17, 87], "user": 4320245} 
click	2	1404166987162	{"section": "jobs", "source": "recommendations"}	{"posting": 17, "user": 4320245} 
rating	3	140416696511	{"rating": 4}	{"posting": 187, "user": 4320245} 

Relationships - Version B (more complex recommendations):

impression	1	1404166899123	{"section": "jobs", "source": "recommendations"}	{"postings": [{"id": 42, "score": 0.67, "reasons": ["mlt", "cf", "haves"]}, {"id": 17, "score": 0.51, "reasons": []}, {"id": 87, "score": 0.14, "reasons": ["cf"]}], "user": 4320245} 
...
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment