UWill Discover

[hosseinfani]

Presenting our findings in the neural team formation at the upcoming undergrad research conference at the university. Here are two useful links. Please plan accordingly.

General information https://www.uwindsor.ca/faculty/recruitment/362/uwill-discover-student-research-conference

Past UWill Discover Conference as examples of student work: https://scholar.uwindsor.ca/uwilldiscover/

[hosseinfani]

Sample posters: https://github.com/fani-lab/ReQue/blob/master/poster.jpg https://github.com/fani-lab/ReQue/blob/master/poster.pub https://github.com/hosseinfani/product_title_quality_cikm_2017/blob/master/Poster.jpg

[dhwanipatel14]

The proposal for Uwill Discover should be submitted by Feb 11, that should include max 250 words describing your research experience and your findings. I am working on it.

The conference will be held on March 4 were I will presenting our research related to Team Formation

[dhwanipatel14]

Hi, @hosseinfani can you provide your suggestions and feedback to this abstract I made draft : In conference they told to consider 4 main questions to write abstarct which are :

1. Why you did this study?
2. What did you do and how?
3. What are your findings?
4. What problem you tried to solve?

Here is draft abstarct : I have worked on research for “Effective Team Formation” which refers to task of forming effective teams of individuals based on their skills and proposed neural machine learning approaches to solve the problem. My main goal to do this research, is to learning machine learning and deep learning and using this learning built efficient model to solve real-world problems. We aim to form a team of experts who collectively hold a set of required skills and can effectively cooperate for the successful completion of the task at hand. Here, we refer to “Team” as the primary vehicle for coordinating experts with diverse skills for particular project, and Team formation (TF) has firsthand effects on creating organizational performance.

In our findings, we found prior works can be distinguished based on their optimization method for finding the optimum team, namely, search-based and learning-based. Search-based approaches search for the optimum team over subgraphs of the expert collaboration network or via integer programming. In contrast, learning-based methods employ machine learning algorithms to learn the distributions of experts and skills from successful teams in the past to draw future teams. The main issue we face is when training data suffers from the long-tail phenomenon where few experts have most of the successful collaborations while the majority has participated sparingly, machine learning models are prone to overfitting. To overcome this issue, we propose an optimization objective that leverages both successful and virtually unsuccessful teams via various negative sampling heuristics. Our experiments on two large-scale benchmark datasets, dblp and imdb, show that neural models that take unsuccessful instances (negative samples) into account are more efficient and effective in training and inference, respectively.

[hosseinfani]

@dhwanipatel14 Thank you for the abstract. It's good. I'll do some revision and will update you. At this time (submission), do we need a poster or just an abstract?

[hosseinfani]

@dhwanipatel14 Automating Team Formation Using Machine Learning Collaborative teams are the primary vehicle for coordinating experts with diverse skills for a particular project in academia, manufacturing, freelancing, and the healthcare sector. Forming a successful team whose members can effectively collaborate and deliver the outcomes within the specified constraints, such as planned budget and timeline, is challenging due to the immense number of candidates with various backgrounds, skills, and personality traits, as well as unknown synergistic balance among them; not all teams with best experts are necessarily successful. Historically, teams have been formed by relying on experience and instinct, resulting in suboptimal team composition due to the incomprehensive knowledge of candidates and hidden cognitive biases, among others.

In our research group, we propose machine learning models that learn relationships among experts and their social attributes through neural architectures to automate forming teams. We aimed at bringing efficiency while maintaining efficacy by employing inherently iterative and online learning procedures in neural architectures. We also aimed at utilizing unsuccessful teams to convey complementary negative signals to neural models. Most real-world datasets, however, do not have explicit unsuccessful teams (e.g., collections of rejected papers.) In the absence of unsuccessful training instances and based on the closed-world assumption, we assume no currently known team of experts for the required skills is to be unsuccessful. Our experiments on two large-scale benchmark datasets, DBLP and IMDB, show that neural models that take unsuccessful teams into account are more efficient and effective in forming collaborative teams.

[dhwanipatel14]

@hosseinfani The abstract looks good and it is under 250 words.

Here is the link were I need to make proposal Submission Link. And official deadline to make this proposal is Feb 18.

Thanks.

[hosseinfani]

@dhwanipatel14 Thanks. The Twitter account and this poster says Feb 11. Can you please submit it by tomorrow to make sure we have something submitted. Later we can resubmit.

[dhwanipatel14]

Hello @hosseinfani , I want to ask shall I put your name under co-author as submission ask for it and faculty sponsor?

[hosseinfani]

Yes, put my name "Hossein Fani"

Please put these co-authors in order: Your name Arman Dashti Karan Saxena Hossein Fani

[dhwanipatel14]

I have made submission proposal .

Thanks

[hosseinfani]

thanks. finger crossrd! please forward any submission email/notification, etc to my email

[hosseinfani]

Our submission got accepted for oral and poster presentation. Congrats to: @dhwanipatel14 @VaghehDashti @karan96