Real-Time Web Scale Event Summarization Using Sequential Decision Making

Authors: Chris Kedzie, Fernando Diaz, Kathleen McKeown

IJCAI 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We demonstrate a 28.3% improvement in summary F1 and a 43.8% improvement in time-sensitive F1 metrics.
Researcher Affiliation Collaboration Chris Kedzie Columbia University Dept. of Computer Science kedzie@cs.columbia.edu Fernando Diaz Microsoft Research fdiaz@microsoft.com Kathleen Mc Keown Columbia University Dept. of Computer Science kathy@cs.columbia.edu
Pseudocode Yes Algorithm 1: Locally optimal learning to search.
Open Source Code Yes source code is available at: https://github.com/kedz/ijcai2016
Open Datasets Yes We evaluate our method on the publicly available TREC Temporal Summarization Track data.2 This data is comprised of three parts. ... 2http://www.trec-ts.org/
Dataset Splits Yes To evaluate our model, we randomly select five events to use as a development set and then perform a leave-one-out style evaluation on the remaining 39 events. In order to avoid over-fitting, we select the model iteration for each training fold based on its performance (in F1 score of expected gain and comprehensiveness) on the development set.
Hardware Specification No The paper does not provide specific hardware details (e.g., GPU/CPU models, memory) used for running experiments.
Software Dependencies No The paper mentions 'WordNet' and 'python-goose' but does not provide version numbers for these or any other software dependencies.
Experiment Setup Yes we downsample each stream to a length of 100 sentences. The downsampling is done uniformly over the entire stream. This is repeated 10 times for each training event to create a total of 380 training streams. The development set was used to set the threshold. The time window size, similarity threshold, and an offset for the cluster preference are tuned on the development set.