search

iLampard / lamp

PyTorch implementation of the paper "Language Models Can Improve Event Prediction by Few-Shot Abductive Reasoning", NeurIPS 2023
Apache License 2.0
47 stars 6 forks source link

Handling Duplicate Time Sequences in GDELT Dataset #5

Closed eogns282 closed 6 months ago

eogns282 commented 6 months ago

Hi,

Thank you for sharing your interesting research. While examining GDELT data you uploaded, I noticed numerous duplicate entries in time sequences. Could you clarify how these duplicates are processed within your TPP models? Understanding your approach to handling these repetitions would be incredibly helpful for my research.

Thank you.

iLampard commented 6 months ago

Hi,

if i remember correctly, duplicate entries are treated as normal input into the model. for details of the computation, you can have a look at the KnowEvolve paper, we simply follow it to compute the intensities.