eRisk, a CLEF lab oriented to early risk prediction on theInternet, started in 2017 as a forum to foster experimentation on earlyrisk detection. After four editions (2017, 2018, 2019 and 2020), the labhas created many reference collections in the field and organized multipleearly risk detection challenges using those datasets. Each challenge focused on a specific early risk detection problem (e.g., depression, anorexiaor self-harm). This paper describes the work done so far, discusses themain lessons learned over the past editions and the plans for the eRisk2021 edition, where we introduced pathological gambling as a new earlyrisk detection challenge.
Keywords: early risk, evaluation, test collections, self-harm, gambling, depression