Phase-annotated conversational data is essential for studying how online grooming of minors progresses and for developing early-warning tools. A new arXiv paper introduces LURE, a synthetic corpus of online-grooming exchanges that includes phase labels, offering a resource for researchers in this area.

The abstract notes that the only publicly available source of such data is the Perverted-Justice archive, though the sentence is cut off in the source, so the specific limitations of that archive are not fully described here. The paper positions LURE as a response to that gap.

Because only the abstract was provided, details on how LURE was generated or validated are not available in the source. The significance lies in the corpus itself: a phase-labelled dataset that could support both analysis of grooming dynamics and the development of automated early-detection systems.