Comments (2)
Title: Fair Team Recommendations for Multidisciplinary Projects
Venue: WI
Year: 2019
Introduction:
Recommender systems could be quite beneficial in diverse team-building challenges for projects. The amount of needed human work can be greatly decreased when there are thousands of candidates and dozens of teams to form. Furthermore, manually maintaining the fairness aspect amongst teams is a difficult chore, which the recommender system algorithm can handle.
This paper focuses on the problem of team recommendation, in which teams must meet multidisciplinary requirements and team members are chosen based on how well their talents match the needs. When building numerous teams, it's also difficult to evenly distribute the best members throughout the teams.
There are two strategies presented in this paper: 1)Brute Force 2)A Faster Heuristic
Main concern: If numerous teams are formed, the top candidates may be assigned to the first team, leaving the less qualified individuals for the other projects. Therefore, to allocate good members to all teams, the fairness element of this team formation should be considered as well.
Dataset: The dataset is a pre-processed version of DBLP that has 7428 lines. Each line corresponds to a researcher and contains the person's name and a varying number of skill tags related to that person. Each researcher has at least one skill and there are 4480 unique skills among all people.
Methods:
Brute Force Algorithm:
From the set of available applicants A, all feasible team combinations T of k members are created for each project p in the set of projects P. The scoreTP function is then used to calculate the score for each possible team t in T in relation to the project p. (t, p). The team with the highest score is selected as the best team for that project, and its members are withdrawn from the pool of candidates A. For each project, the process of forming a team is repeated.
Heuristic Algorithm:
Due to the computation of all possible team combinations, the brute force approach is computationally expensive. As a result, a heuristic is proposed that can be used to reduce calculations while maintaining suggestion efficacy.
Rather than producing all possible team combinations, our heuristic generates scoreAP(a, p) between each applicant an in the set A and each project p in the set of projects P. The candidate a with the highest scoreAP for a specific project p is then selected as a member of that project team and removed from the pool of applicants A. This procedure is done k times until the project p team has selected all of its k members. The procedure of forming teams is then repeated for all of the other projects in the set P.
Future Work:
Defining and adding new aspects of fairness to it, so that we can form more productive and diverse teams at the same time.
from adila.
@Rounique
Nice summary.
It lacks some parts though: gaps of previous work, gaps of this work, code, and evaluation
More importantly, where the fairness is applied in the proposed two methods!
from adila.
Related Issues (20)
- The Epic Road to Fairness HOT 7
- Zero before/after NDKL for every input after code refactor HOT 7
- Summary of an Unsuccessful Attempt with Double-cutoff Method HOT 4
- equality of opportunity fairness criteria HOT 6
- Removing torch from Adila's pipeline HOT 5
- Key Error While Reranking with det_cons HOT 2
- 2017:CIKM: FA*IR: A Fair Top-k Ranking Algorithm HOT 4
- Bubble Plot for Fairness vs Utility HOT 10
- Cutoff for popular/nonpopular based on distribution diagram HOT 17
- IR Metrics Without a Cutoff HOT 2
- 2017:SSDBM:Measuring Fairness in Ranked Outputs
- Investigating gender fairness of recommendation algorithms in the music domain HOT 3
- pipeline progress flow HOT 2
- 2022:ACM: Fairness in Ranking, Part II: Learning-to-Rank and Recommender Systems (UNDER CONSTRUCTION)
- FA*IR Method for Neural Team Formation HOT 7
- 2022:KDD:Fair Ranking as Fair Division: Impact-Based Individual Fairness in Ranking
- Discussion about Adding a Baseline to Adila HOT 7
- uspt gender labels and experiments HOT 4
- Gender Distribution For imdb, dblp and uspt Datasets.
- Exposure Metric
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from adila.