Selecting Sequences of Items via Submodular Maximization

Authors: Sebastian Tschiatschek, Adish Singla, Andreas Krause

AAAI 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We demonstrate the effectiveness of our algorithm in synthetic and real world experiments on a movie recommendation dataset.
Researcher Affiliation Academia Sebastian Tschiatschek ETH Zurich sebastian.tschiatschek@inf.ethz.ch Adish Singla ETH Zurich adish.singla@inf.ethz.ch Andreas Krause ETH Zurich krausea@ethz.ch
Pseudocode Yes Algorithm 1 GREEDY: Node-based Greedy Algorithm; Algorithm 2 OMEGA: Edge-based Greedy Algorithm with Reordering; Algorithm 3 REORDER: Compute Sequence of Items from Set of Edges
Open Source Code No The paper refers to an 'extended version' for proofs, but does not provide any specific statement or link for the availability of source code for the described methodology.
Open Datasets Yes We performed real world movie recommendation experiments on the Movielens 1M dataset2. This dataset contains 1,000,209 ratings of 6,040 users for 3,706 movies. [...] 2http://grouplens.org/datasets/movielens/1m/
Dataset Splits No The paper states, 'We randomly partitioned the data D into training data Dtrain and testing data Dtest such that |Dtest| = 500,' but does not specify details about a validation split or other specific partitioning methodologies.
Hardware Specification No The paper does not provide specific hardware details such as GPU/CPU models or memory used for running its experiments.
Software Dependencies No The paper does not provide specific ancillary software details with version numbers (e.g., library or solver names with versions).
Experiment Setup Yes We model dependencies between the last z items in σpr and the items that can be selected, where we used z {1, 2, 5, } in our experiments (z = means that all items in σpr are considered).