Recognition Of Textual Entailment Research Articles

Adversarial shared-private model for cross-domain clinical text entailment recognition

The recognition of textual entailment (RTE) as the main text understanding task is crucial to the application in biomedical and clinical field, however, the developing of which has been hindered, due to the scarcity of the data annotation. In this work, we propose a domain adaptation framework for the cross-domain clinical RTE. We first construct a hierarchical feature encoder architecture for fully exploring the interactions between the input sentence pair. We then establish shared and private feature extractors based on the feature encoder, for capturing both the domain-specific and domain-invariant features. We further introduce a domain discriminator with the adversarial training algorithm for enhancing the cross-domain transferring. Based on the real-world Chinese dataset, our framework achieves significantly enhanced performances against baseline domain adaptation methods, on the few-shot and zero-shot transferring settings. Further analysis reveals that our model is effective for the cross-domain clinical RTE.

Recognition of Partial Textual Entailment for Indian Social Media Text

Textual entailment (TE) is a unidirectional relationship between two expressions where the meaning of one expression called Hypothesis (H), infers from the other expression called Text (T). The definition of TE is rigid in a sense that if the H entails from T but lacks minor information or have some additional information, then the pair treated as non-entailed. In such cases we could not measure the relatedness of a T-H pair. Partial textual entailment (PTE) is a possible solution of this problem which defines partial entailment relation between a T-H pair. PTE relationship can plays an important role in different Natural Language Processing (NLP) applications like text summarization and question-answering system by reducing redundant information. In this paper we investigate the idea of PTE for Indian social media text (SMT). We developed a PTE annotated corpus for Bengali tweets and proposed a Sequential Minimal Optimization (SMO) based PTE recognition approach. We also evaluated our proposed approach through experiment results.

Different Models and Approaches of Textual Entailment Recognition

Variability of semantic expression is a fundamental phenomenon of a natural language where same meaning can be expressed by different texts. The process of inferring a text from another is called textual entailment. Textual Entailment is useful in a wide range of applications, including question answering, summarization, text generation, and machine translation. The recognition of textual entailment is one of the recent challenges of the Natural Language Processing (NLP) domain. This paper summarizes key ideas from the area of textual entailment recognition by considering in turn the different recognition models. The paper points to prominent testing data, training data, resources and Performance Evaluation for each model. Also this paper compares between textual entailment models according to the method which used, the result of each method and the strong and weakness of each method.

3.1.2 Machine Learning technologies for the Requirements Analysis in Complex Systems

AbstractRequirement Analysis is a relevant application area for a variety of Semantic Technologies related to the extraction, disambiguation and exploitation of the knowledge derived from technical requirement documents. Most methods rely on shallow language processing technologies for the automatic extraction of core concepts (e.g. components/devices, their parts and functionalities) and norms (e.g. constraints on the use of components). Few works have been devoted to study paraphrasing (i.e. textual equivalence between requirement definitions) for consistency checking and redundancy elimination. We propose here a distributional method to train a kernel‐based learning algorithm (i.e. SVM), as a cost‐effective approach to validate requirements from text in support of Requirement Analysis in the design of a Complex Systems, i.e. Naval Combat Systems. These latter are complex systems based on software components able to manage all the Combat System Equipment in different mission scenarios. We will describe the application of Recognition of Textual Entailment (RTE) techniques based on data‐driven learning methods to this scenario. While modeling the asymmetric logical relation of entailment between two textual descriptions (i.e. an hypothesis and its thesis), RTE can be here applied to the validation step of compositions between system functionalities described in the requirement specification texts. Early evidences are here discussed, as a proof of the strong applicability of the method to the general case and to other realistic scenarios in System Engineering.