Neural Topic Modeling with Continual Lifelong Learning

Authors: Pankaj Gupta, Yatin Chaudhary, Thomas Runkler, Hinrich Schuetze

ICML 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Intensive experimental results show improved topic modeling on future task while retaining past learning, quantified by information retrieval, topic coherence and generalization capabilities.
Researcher Affiliation Collaboration 1Corporate Technology, Siemens AG Munich, Germany 2CIS, University of Munich (LMU) Munich, Germany.
Pseudocode Yes Algorithm 1 Lifelong Neural Topic Modeling using Doc NADE
Open Source Code Yes Code: https://github.com/pgcool/ Lifelong-Neural-Topic-Modeling
Open Datasets Yes AGnews, TMN, R21578 and 20NS (20News Groups), and three short-text (low-resource) corpora ΩT +1 as future tasks T + 1: 20NSshort, TMNtitle and R21578title.
Dataset Splits No The paper states 'PPL (Algorithm 2: line #10) is used for model selection and adjusting parameters Θt and hyperparameters Φt.' and refers to a 'test set', but does not explicitly provide specific train/validation/test dataset split percentages or sample counts for a distinct validation set.
Hardware Specification Yes Figures 3, 4 and 5 show average runtime (r-time) for each training epoch of different LNTM approaches, run on an NVIDIA Tesla K80 Processor (RAM: 12 GB) to a maximum of 100 epochs.
Software Dependencies No The paper does not specify software dependencies with version numbers (e.g., Python version, library versions like TensorFlow or PyTorch).
Experiment Setup Yes Figures 3, 4 and 5 show average runtime (r-time) for each training epoch of different LNTM approaches, run on an NVIDIA Tesla K80 Processor (RAM: 12 GB) to a maximum of 100 epochs.