xai-org / xai-org/x-algorithm

Personalized For You Feed for Final-Year Students and Job Seekers

Open
#42 2 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Rust
Stars
33.3k
Forks
5.4k
Avg merge
1d 17h
Merged PRs (30d)
1

Description

Personalized For You Feed for Final-Year Students and Job Seekers

Problem

Students in their final year and users actively searching for jobs need a For You timeline that prioritizes content useful for their employment search.

The current recommendation approach is primarily optimized around engagement signals such as likes, replies, reposts, shares, clicks, and dwell time. This can result in a feed containing a large amount of general or entertainment content even when the user's primary objective is finding a job.

Goal

Introduce a personalized recommendation mode for users who are:

  • In their final year of study
  • Recently graduated
  • Actively searching for jobs
  • Interested in entry-level or fresher opportunities

For these users, approximately 80% of the For You timeline should contain hiring and employment-related posts.

The remaining 20% can contain normal personalized content so that the timeline does not become completely one-dimensional.

Proposed Feed Distribution

For You Timeline

┌─────────────────────────────────────────┐
│ 80% Hiring / Career Opportunities       │
│                                         │
│ • Job openings                          │
│ • Fresher hiring                        │
│ • Entry-level roles                     │
│ • Graduate programs                     │
│ • Internships                           │
│ • Referral opportunities                │
│ • Walk-in interviews                    │
│ • Recruitment announcements             │
│ • Hiring threads                        │
│ • Company career opportunities          │
├─────────────────────────────────────────┤
│ 20% General Personalized Content        │
│                                         │
│ • Technology                            │
│ • AI / programming                     │
│ • Career advice                         │
│ • Industry discussions                  │
│ • Other user interests                  │
└─────────────────────────────────────────┘

Recommendation Requirements

The ranking system should identify and score hiring-related posts using signals such as the following:

  • Hiring intent
  • Entry-level / fresher relevance
  • Job role relevance
  • Location relevance
  • Recency
  • Company / recruiter relevance
  • Internship or graduate-program relevance
  • Referral intent

A post such as:

"Hiring fresh graduates for Python Developer roles in Particular Place"

should receive a substantially higher ranking than a general post from the same topic area when the user has an active entry-level job-search preference.

Personalization

The 80% target should be treated as a personalized feed objective, rather than a permanent rule applied to every user.

The system should activate this behavior when the user explicitly indicates that they are:

  • A final-year student
  • A recent graduate
  • Looking for a job
  • Interested in entry-level opportunities

User interactions should then provide additional signals to improve ranking over time.

For example:

User actions
    ↓
Likes / Bookmarks / Opens / Dwell Time
    ↓
Hiring-interest profile
    ↓
Personalized ranking
    ↓
Improved For You timeline

Acceptance Criteria

  • A user can indicate that they are a final-year student or active job seeker.
  • The recommendation system can classify posts according to hiring relevance.
  • Hiring-related posts receive a strong ranking boost for eligible users.
  • The resulting For You timeline targets approximately 80% hiring-related content over a sufficiently large feed sample.
  • Entry-level and fresher opportunities receive higher priority than senior-level opportunities for entry-level users.
  • Location and role preferences can influence the ranking.
  • Non-hiring content is still allowed within the remaining portion of the feed.
  • The system continues learning from user interactions and adjusts recommendations over time.
  • Feed diversity is preserved so that the 80% hiring content does not become repetitive or dominated by a single author/company.
  • The ranking system should measure the actual percentage of hiring-related posts delivered so the 80% target can be evaluated objectively.

Success Metric

The primary metric for this feature should be the following:

Hiring Content Ratio = Hiring-related posts shown / Total For You posts shown

Target:

≈ 80% for users who have explicitly enabled the job-search personalization mode.

Secondary metrics can include:

  • Relevant job-post click rate
  • Job-post bookmark rate
  • Recruiter/company profile visits
  • Application-intent interactions
  • User satisfaction with recommendations
  • Repetition/diversity rate

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

No files, tests, or entry points are named. Start by locating the existing For You ranking, user-preference, and post-classification components, then determine how hiring relevance and feed composition are measured. Done means eligible users can enable the mode, receive an approximately 80% hiring-related feed with diversity, and have the ratio and secondary metrics evaluated.

Written by the indexing model from the issue text.

Assessment

Tech stack
rust
Domain
backend, machine-learning
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Quiet
Clarity
Mostly clear
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.