Knowledge-Based Semantic Embedding for Machine Translation

Abstract

In this paper, with the help of knowledge base, we build and formulate a semantic space to connect the source and target languages, and apply it to the sequence-to-sequence framework to propose a Knowledge-Based Semantic Embedding (KBSE) method. In our KB-SE method, the source sentence is firstly mapped into a knowledge based semantic space, and the target sentence is generated using a recurrent neural network with the internal meaning preserved. Experiments are conducted on two translation tasks, the electric business data and movie data, and the results show that our proposed method can achieve outstanding performance, compared with both the traditional SMT methods and the existing encoder-decoder models.

Syntax for Semantic Role Labeling, To Be, Or Not To Be

“…Specifically, we seek to identify arguments and label their semantic roles given a predicate. SRL is an impor-tant method to obtain semantic information beneficial to a wide range of natural language processing (NLP) tasks, including machine translation (Shi et al, 2016), question answering (Berant et al, 2013;Yih et al, 2016) and discourse relation sense classification (Mihaylov and Frank, 2016).…”

Neural Natural Language Inference Models Enhanced with External Knowledge

“…In general, external knowledge has shown to be effective in neural networks for other NLP tasks, including word embedding (Chen et al, 2015;Faruqui et al, 2015;Liu et al, 2015;Wieting et al, 2015;Mrksic et al, 2017), machine translation (Shi et al, 2016;Zhang et al, 2017b), language modeling (Ahn et al, 2016), and dialogue systems .…”

Machine Translation using Semantic Web Technologies: A Survey

“…[130] 2016 SMT SPARQL Ontologies Human N. Abdulaziz et al [131] 2016 SMT SPARQL Ontologies Human J. Du et al [132] 2016 SMT SPARQL LOD Automatic A. Srivastava et al [133] 2016 SMT SPARQL + Annotation LOD Automatic C. Shi et al [134] 2016 NMT Annotation LOD Automatic + Human A. Srivastava et al [135] 2017 SMT SPARQL + Annotation LOD Automatic of ambiguity in USL. The crucial contribution according to authors was the creation of a new domain-specific language based on the ontologies which may be further used for editing and processing future works in the translation of USL using ontologies.…”

“…• C. Shi et al [134] built a semantic embedding model relying on knowledge-bases to be used in NMT systems. The work is dubbed Knowledge Base Semantic Embedding (KBSE), which consists of mapping a source sentence to triples and then using these triples to extract the internal meaning of words to generate a target sentence.…”