Combining Axiom Injection and Knowledge Base Completion for Efficient Natural Language Inference

Yoshikawa, Masashi; Mineshima, Koji; Noji, Hiroshi; Bekki, Daisuke

doi:10.1609/aaai.v33i01.33017410

Cited by 7 publications

(5 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In future work, we will extend our analysis to cover the more complex constructions mentioned in Section 3. We are also considering combining our system with an abduction mechanism that uses large knowledge bases (Yoshikawa et al, 2019) for handling commonsense reasoning with external knowledge.…”

Section: Discussionmentioning

confidence: 99%

Logical Inferences with Comparatives and Generalized Quantifiers

Haruta

Mineshima

Bekki

2020

Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics: Student Research Workshop

Self Cite

View full text Add to dashboard Cite

Comparative constructions pose a challenge in Natural Language Inference (NLI), which is the task of determining whether a text entails a hypothesis. Comparatives are structurally complex in that they interact with other linguistic phenomena such as quantifiers, numerals, and lexical antonyms. In formal semantics, there is a rich body of work on comparatives and gradable expressions using the notion of degree. However, a logical inference system for comparatives has not been sufficiently developed for use in the NLI task. In this paper, we present a compositional semantics that maps various comparative constructions in English to semantic representations via Combinatory Categorial Grammar (CCG) parsers and combine it with an inference system based on automated theorem proving. We evaluate our system on three NLI datasets that contain complex logical inferences with comparatives, generalized quantifiers, and numerals. We show that the system outperforms previous logic-based systems as well as recent deep learning-based models.

show abstract

Section: Discussionmentioning

confidence: 99%

Logical Inferences with Comparatives and Generalized Quantifiers

Haruta

Mineshima

Bekki

2020

Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics: Student Research Workshop

Self Cite

View full text Add to dashboard Cite

show abstract

“…Yoshikawa et al [6] proposed an alternative method for extending ccg2lambda by introducing an efficient mechanism for axiom injection based on Knowledge Base Completion (KBC) models [39]. KBC models are machine learning models that have recently seen significant advancements.…”

Section: Related Work a Logic-based Approaches To Rte Tasksmentioning

confidence: 99%

“…This task predicts whether a given premise sentence entails a hypothesis sentence. Logic-based approaches [2], [3], [4], [5], [6] and machine learning approaches [7], [8], [9] are the primary approaches to RTE. Logic-based approaches use logical formulas to represent the linguistic meanings of sentences and try to prove entailment relations between formulas.…”

Section: Introductionmentioning

confidence: 99%

Logic-Based Inference With Phrase Abduction Using Vision-and-Language Models

Tomihari,

Yanaka

2023

IEEE Access

View full text Add to dashboard Cite

Recognizing Textual Entailment (RTE) is among the most fundamental tasks in natural language processing applications, such as question answering and machine translation. One of the main challenges in logic-based approaches to this task is the lack of background knowledge. This study proposes a logical inference system with phrasal knowledge by comparing their visual representations based on the intuition that visual representations enable people to judge entailment relations. First, we obtain candidate phrase pairs for phrasal knowledge from logical inference. Second, using a vision-and-language model, we acquire the visual representations of these phrases in the form of images or embedding vectors. Finally, we compare these obtained visual representations to determine whether to inject the knowledge corresponding to the candidate. In addition to simple similarity between phrases, we also consider asymmetric relations when comparing visual representations. Our logical inference system improved accuracy on the SICK dataset compared with a previous logical inference system, SPSA (Selector of Predicates with Shared Arguments). Moreover, our asymmetric evaluation functions using vision-and-language models are effective at capturing the entailment relations of word pairs in HyperLex.INDEX TERMS Natural language processing, recognizing textual entailment, vision and language.

show abstract

“…The method proposed by Shen et al [33] uses a global memory and a controller module to learn multi-hop paths in vector space and infers missing facts jointly without any human-designed procedure. Yoshikawa et al [43] achieve KBC by combining axiom injection as an inference basis. By constructing rich features, Komninos and Manandhar [47] use a neural network to complete missing relations.…”

Section: Knowledge Base Completionmentioning

confidence: 99%