Recognizing Textual Entailment by Soft Dependency Tree Matching
We present a rule - based method for recognizing entailmen t relation between a pair of text fragments by comparing their dependency tree structures. We used a dependency parser to generate the dependency triple s of the text – hypothesis pair s . A dependency triple is a n arc in the dependency par...
| Autores: | , , , |
|---|---|
| Tipo de recurso: | artículo |
| Estado: | Versión publicada |
| Fecha de publicación: | 2015 |
| País: | México |
| Institución: | Instituto Politécnico Nacional |
| Repositorio: | Redalyc-IPN |
| OAI Identifier: | oai:redalyc.org:61543181006 |
| Acceso en línea: | https://www.redalyc.org/articulo.oa?id=61543181006 |
| Access Level: | acceso abierto |
| Palabra clave: | Computación rules PETE dataset Textual entailment dependency parsing dependency relation matching |
| Sumario: | We present a rule - based method for recognizing entailmen t relation between a pair of text fragments by comparing their dependency tree structures. We used a dependency parser to generate the dependency triple s of the text – hypothesis pair s . A dependency triple is a n arc in the dependency parse tree. Each triple in the hypothesis is checked against all the triple s in the text to find a matching pair . We have developed a number of matching rules after a detailed analysis of the PETE dataset , which we used for the experiments . A successful match satisfying any of th ese rules assign s a matching score of 1 to the child node of th at particular arc in the hypothesis dependency tree. Then the dependency parse tree is traversed in post - order way to obtain the final entai lment score at the root node. The score s of the leaf nodes are propagated from the bottom of the tree to the non - leaf nodes , up to the root node. The entailment score of the root node is compared against a predefined threshold value to make the entailment decision . Experimental result s on the PETE dataset sh ow an accuracy of 87.69 % on the dev elopment set and 73. 75 % on the test set , which outperforms the state - of - the - art results reported on this dataset so far . We did not use any other NLP tools or knowledge sources , to emphasize the role of dependency parsing in rec ognizing textual entailment. |
|---|