Joint Posterior Revision of NLP Annotations via Ontological Knowledge
Authors: Marco Rospocher, Francesco Corcoglioniti
IJCAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | In a concrete scenario with two stateof-the-art tools for NERC and EL, we experimentally show on three reference datasets that for these tasks, the joint annotation revision performed by the model consistently improves on the original results of the tools. |
| Researcher Affiliation | Academia | Marco Rospocher and Francesco Corcoglioniti Fondazione Bruno Kessler (FBK-irst) {rospocher, corcoglio}@fbk.eu |
| Pseudocode | No | The paper presents mathematical formulations and derivations but does not include structured pseudocode or algorithm blocks. |
| Open Source Code | No | The paper states JPARK was implemented as a Java module of PIKES, an open-source framework, but does not explicitly state that the JPARK module itself is open-source or provide a direct link to the specific JPARK implementation used in the paper's experiments. |
| Open Datasets | Yes | AIDA Co NLL-YAGO [Hoffart et al., 2011] This dataset consists of 1,393 English news wire articles from Reuters, with 34,999 mentions hand-annotated with named entity types... MEANTIME [Minard et al., 2016]... TAC-KBP [Ji et al., 2011]... |
| Dataset Splits | Yes | We use AIDA eng.train as the gold standard G for estimating the probabilities P(C|a NERC, K)... and we use AIDA eng.testa to optimize the model hyperparameters of Section 3, i.e., n (best value = 1000, corresponding to 54 YAGO classes and 2041 class sets) and α (best value = 0.02). The evaluation is separately conducted on three datasets: AIDA eng.testb, MEANTIME and TACKBP. |
| 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 software like Stanford NER, DBpedia Spotlight, and PIKES, but does not provide specific version numbers for any of them. |
| Experiment Setup | No | The paper mentions that hyperparameters 'n' and 'α' were optimized, but it does not provide specific experimental setup details such as learning rates, batch sizes, or optimizer settings for training the NLP tools themselves or the JPARK model. |