Skip to content

ML-power of Randorm: Recommendations and Allocation Algorithms

License

Notifications You must be signed in to change notification settings

randorm/research

Repository files navigation

Data graph

In preparation for the check-in of Innopolis University applicants at InnoBootCamp 2023, we collected a data graph showing the relationship between participants, their answers to the questions from the feed. We have also included the result of the applicant distribution by room in this graph.

The graph is presented as three files:

Below you will find a description of each node stored in the data graph.

Participant

This node represents a Randorm user.

Model

Field Type Required Description
id Integer Yes Unique participant identifier.
created_at Integer Yes Date the user was registered. The date is in "days until deadline" format.
gender String Yes User gender. Can be either male or female.
subscriber_count Integer Yes Number of subscribers. Note that the value may be higher than the number of identifiers in subscriber_ids.
subscriber_ids Array of Integer Yes Set of unique subscriber identifiers. Note that only users who meet the distribution requirements are presented.
subscription_count Integer Yes Number of subscriptions. Note that the value may be higher than the number of identifiers in subscription_ids.
subscription_ids Array of Integer Yes Set of unique subscription identifiers. Note that only users who meet the distribution requirements are presented.
viewed_count Integer Yes Number of viewed users. Note that the value may be higher than the number of identifiers in viewed_ids.
viewed_ids Array of Integer Yes Set of unique viewed user identifiers. Note that only users who meet the distribution requirements are presented.
views Integer Yes Number of views by other users. Note that users may have looked more than once.
roommate_ids Array of Integer Yes Set of unique roommate identifiers.

Note that participant is anonymized, i.e. participant identifier does NOT match the real user identifier in the system.

Example

{
  "id": 42,
  "created_at": 12,
  "gender": "male",
  "subscriber_count": 14,
  "subscriber_ids": [
    169,
    65,
    120
    // ...
  ],
  "subscription_count": 4,
  "subscription_ids": [
    169,
    65,
    120
    // ...
  ],
  "viewed_count": 4,
  "viewed_ids": [
    169,
    65,
    120
    // ...
  ],
  "views": 59,
  "roommate_ids": [
    128,
    46,
    120
  ]
}

Field

This node represents a single-choice question.

Model

Field Type Required Description
id Integer Yes Unique field identifier.
question String Yes Question.
options Array of String Yes Set of options to choose.

Note that you are only given single-choice questions. No text or multiple-choice questions are provided.

Example

{
  "id": 3,
  "question": "I'd like a roommate...",
  "options": [
    "Similar to my preferences",
    "Opposite to my preferences"
  ]
}

Answer

This node represents a participant's answer to a single-choice question.

Model

Field Type Required Description
field_id Integer Yes Unique field identifier.
respondent_id Integer Yes Unique participant identifier.
option Integer Yes Chosen option index.

Note that some users did not answer the questions because they joined late.

Example

{
  "field_id": 3,
  "respondent_id": 42,
  "option": 0
}

Git LFS

This repository is using Git LFS (Git Large File Storage) system. Make sure you have installed the corresponding extension to your git enviroment. Installation Guide

About

ML-power of Randorm: Recommendations and Allocation Algorithms

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published