TalentCLEF at CLEF2025: Skill and Job Title Intelligence for Human Capital Management
Abstract
TalentCLEF introduces an innovative evaluation lab designed to foster the development of Natural Language Processing systems in the field of Human Capital Management. In a corporate environment that is rapidly moving towards a globalized and multilingual workforce, organizations are increasingly relying on language technologies to optimize and accelerate recruitment processes. This scenario brings a critical challenge: developing systems that guarantee fairness in outcomes with the ability to function across multiple languages and adapt to various industries. TalentCLEF addresses this challenge by creating a public benchmark that promotes both the development and evaluation of NLP systems in this area, carefully considering these aspects in its design. The first edition of TalentCLEF will be held at CLEF 2025 and will have two tasks: (i) Task A - Multilingual Job Title Matching, where the goal is to develop systems capable of retrieving job positions similar to a given one; and (ii) Task B - Job Title-Based Skill Prediction, where teams will focus on creating systems that identify professional skills relevant to a specific job position.
