Speaker
Description
Flavor tagging is a key requirement for the physics program of future e+e- colliders such as FCC-ee, where precision measurements of Higgs, electroweak, and flavor physics observables rely on the ability to identify quark flavors in hadronic final states. While charm and bottom tagging have been extensively studied, strange quark tagging remains comparatively underdeveloped despite its importance for probing the Higgs-strange coupling and related electroweak processes.
In this study, we investigate strange flavor tagging performance using the ILD (International Large Detector) concept adapted for FCC-ee running conditions. Strange jets are characterized by the presence of strange hadrons, in particular K-short and Lambda baryons, which decay via displaced V0 vertices. We examine how the reconstruction quality and selection efficiency of V0 candidates affect the discriminating power for strange-quark identification, using full detector simulation and event reconstruction within the ILD software framework.
Building on this analysis, we develop a machine learning-based flavor tagger that incorporates V0-related observables—such as V0 invariant mass, displaced vertex position, and decay kinematics—alongside standard jet-level and particle flow information. The performance of this tagger is evaluated in terms of strange-tagging efficiency and purity, and compared against baseline approaches that do not explicitly exploit V0 information.
Our results demonstrate the sensitivity of strange tagging performance to V0 reconstruction quality and highlight the potential of ML-based taggers to improve strange-quark identification at future Higgs factories. These findings contribute to ongoing efforts to optimize detector design and reconstruction algorithms for the ILD concept at FCC-ee.