Narrative Planning Model Acquisition from Text Summaries and Descriptions
Authors: Thomas Hayton, Julie Porteous, Joao Ferreira, Alan Lindsay1709-1716
AAAI 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | The aim of the evaluation was to assess the performance of our approach on: (i) pronoun coreference resolution with multiple pronoun references across sentences; (ii) identification of narrative actions and associated objects from multi-clause sentences; and (iii) generating useful baseline narrative planning models, as assessed by expert users. Figure 4 shows system performance on identification of narrative actions 3 and associated objects 4 from the input NL text synopses. The table lists the number of input sentences (S), and then the number of correctly identified Narrative Actions and Objects, against the total number of actions and objects in the input sentences, and the number of errors (where errors are actions and objects which not judged as such in the text). |
| Researcher Affiliation | Academia | Thomas Hayton,1 Julie Porteous,2 Jo ao F. Ferreira,3 Alan Lindsay4 1Teesside University, Middlesbrough, United Kingdom, t.hayton@tees.ac.uk 2School of Science, RMIT University, Melbourne, Australia, julie.porteous@rmit.edu.au 3INESC-ID & Instituto Superior T ecnico, Universidade de Lisboa, Portugal, joao@joaoff.com 4University of Huddersfield, United Kingdom, a.lindsay@hud.ac.uk |
| Pseudocode | Yes | Algorithm 1: Pronoun Coreference Algorithm |
| Open Source Code | No | The paper states 'Our approach is fully implemented in a prototype system' but does not provide a link to its open-source code or explicitly state its release. |
| Open Datasets | Yes | For the evaluation we used the following synopses: Scooby Doo, Friends, House, Jungle Book, Toy Story (TS), Titanic, Merchant of Venice (Mo V), Christmas Carol (CC), Lord of the Flies (Lo F) and The Odyssey (OD) [Synopses]. These were chosen because they are publicly available online resources... The synopses used in the study are available for download from: http://tiny.cc/3o76bz. |
| Dataset Splits | No | The paper lists the synopses used for evaluation (e.g., Scooby Doo, Jungle Book) but does not specify any training, validation, or test dataset splits or cross-validation setup. |
| Hardware Specification | No | The paper does not provide specific hardware details (e.g., GPU/CPU models, memory) used for running its experiments. |
| Software Dependencies | No | The paper mentions 'Stanford Core NLP' and 'spaCy' as tools, and 'Spa Cy 2.1+ 2019' in a citation, but does not list specific version numbers for its own software dependencies. |
| Experiment Setup | No | The paper describes its method in detail, including rules for pronoun coreference, but does not provide specific experimental setup details such as hyperparameter values, training configurations, or system-level settings. |