A Discriminative Approach to Grounded Spoken Language Understanding in Interactive Robotics

Authors: Emanuele Bastianelli, Danilo Croce, Andrea Vanzo, Roberto Basili, Daniele Nardi

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

Reproducibility Variable Result LLM Response
Research Type Experimental Empirical results achieve up to a 40% of relative error reduction.
Researcher Affiliation Academia 1DICII, 2DII, University of Rome Tor Vergata, Rome, Italy 3DIAG, Sapienza University of Rome, Rome, Italy bastianelli@ing.uniroma2.it, {croce,basili}@info.uniroma2.it, {vanzo,nardi}@dis.uniroma1.it
Pseudocode No The paper describes the steps of the proposed approach in natural language text but does not include any explicit pseudocode or algorithm blocks.
Open Source Code No The paper states 'The SVMhmm algorithm has been implemented within the Ke LP framework [Filice et al., 2015].' This refers to a third-party framework and does not indicate that the authors' own implementation for the described methodology is open source.
Open Datasets Yes The evaluation is carried out using the Human-Robot Interaction Corpus (Hu RIC, [Bastianelli et al., 2014b]) a collection of utterances annotated with semantic predicates and paired with (possibly multiple) audio files.
Dataset Splits Yes Measures have been carried out on four tasks, all according to a 5-fold evaluation schema.
Hardware Specification No The paper does not provide specific hardware details (e.g., exact GPU/CPU models, memory amounts) used for running its experiments.
Software Dependencies No The paper mentions software like the 'Ke LP framework' and 'Core NLP parser' but does not provide specific version numbers for these or any other ancillary software components.
Experiment Setup Yes By applying the settings min-count=50, window=5, iter=10 and negative=10 onto the Uk Wa C corpus we derived 250 dimensional word vectors for more than 110, 000 words. The SVMhmm algorithm has been implemented within the Ke LP framework [Filice et al., 2015].