Recommendation Systems

Recommenders are recommendation systems that suggest items of interest for users. The information overload demands effective approaches to filter relevant information for users. Such approaches should effectively predict user ratings for items to maximaze the user consumption of items. Our goal is to propose effective recommenders for differents tasks, investigating their behavior in a real world scenario by contrasting them with the state-of-the-art recommenders reported in literature.

Started in 2018, with no funding.

Prediction on Sports

Sport analytics investigates problems related to sports by exploiting data and sports statistics. Our goal is to propose and evaluate approaches based on machine learning and information retrieval algorithms to predict the occurrence of patterns in events and sports activities.

Started in 2018, with funding from CNPq (grant 2018/22188).

Text Representation

Stemming is the process of reducing inflected words to their root form, the stem. Search engines have been using it to improve the user search experience. Stemming algorithms (stemmers) usually adopt suffix stripping strategies to decide which reductions should be performed over words. Our goal is to propose effective stemmers and investigate their behavior in a real world scenario by contrasting them with the state-of-the-art stemmers reported in literature.

Started in 2016, with funding from PUC Minas (grant FIP 2016/11086-2).

Prediction on Workforce

Workforce analytics, people analytics and HRM analytics refer to the analysis and use of employee data as features for workforce-related planning and decision making. Our goal is to propose and evaluate algorithms, prediction models and intelligent systems for workforce analysis and performance, particularly for attracting, developing and retaining talent and managing human resources.

Started in 2021, with funding from Sólides.