Original Sentence Research Articles

Research on Chinese Sentence Compression Automatic Evaluation Method based on Sentence Hierarchical Information Extraction

People often feel incomprehensible and difficult to understand in the face of huge amounts of information. They instinctively produce the natural need to condense information to save time. Sentence compression technology has been born. The research on sentence compression evaluation methods has developed with the rise of sentence compression technology, and has gradually become a hot issue for sentence compression tasks. The method of sentence compression evaluation is to judge a good or bad cornerstone of compression results. influences. Under the current practical background, the performance evaluation of sentence compression mainly includes two methods: manual evaluation and automatic evaluation. The main problems encountered are: First, most of the research work uses manual evaluation indicators. The compression results under manual evaluation have high accuracy and meet the requirements. People’s daily language habits have the disadvantages of inflexibility and poor reusability. Second, there are few automatic evaluation methods for Chinese sentences, and the existing traditional automatic evaluation methods only consider the morphological similarity between the compressed sentence and the original sentence, and cannot well represent the compression effect. Based on the above practical problems, this paper proposes an automatic evaluation method for Chinese sentence compression based on sentence hierarchical information extraction. Compared with the manual evaluation method and the traditional automatic evaluation method, the Chinese sentence compression automatic evaluation method based on sentence hierarchical information extraction overcomes the disadvantages of high manual evaluation cost and low reusability. It uses the similarity of sentence hierarchical information to diversify. Ground characterization and evaluation of compression effects have significantly improved the rationality and accuracy of traditional automatic evaluation methods Introduction.

Source Attribution: Recovering the Press Releases Behind Health Science News

We explore the task of intrinsic source attribution: inferring which portions of a derived document were adapted from an unobserved source document. Specifically, we model the relationship between news articles and their press release sources using a dataset of 64,784 health science news articles and 23,068 press releases. We approach the problem at the sentence level and work with science journalism professors to develop a four point Likert scale describing the extent to which a news article sentence is derived from the content in the corresponding press release. Because manual annotation of news article - press release pairs is time-consuming, we turn to a mix of expert, non-expert, and heuristic-based annotation to label our dataset. After a small pilot study, which found that humans, when only able to view the text of the news article, struggle to identify which content is derived or not, we compare four different sentence regression models on the task. We find that modeling a sentence's context in the entire document is important, with the best performing model, a sequence regression model with BERT token representations, achieving a spearman's ρ of 0.49 and NDCG@1 of 0.60 on the expert-labeled test set. Examining the model's predictions, we find that it successfully identifies copied or closely paraphrased sentences in articles with a mix of derived and original content, but struggles to differentiate between loosely paraphrased and original sentences in articles with mostly original writing.

Fact-Aware Sentence Split and Rephrase with Permutation Invariant Training

Sentence Split and Rephrase aims to break down a complex sentence into several simple sentences with its meaning preserved. Previous studies tend to address the issue by seq2seq learning from parallel sentence pairs, which takes a complex sentence as input and sequentially generates a series of simple sentences. However, the conventional seq2seq learning has two limitations for this task: (1) it does not take into account the facts stated in the long sentence; As a result, the generated simple sentences may miss or inaccurately state the facts in the original sentence. (2) The order variance of the simple sentences to be generated may confuse the seq2seq model during training because the simple sentences derived from the long source sentence could be in any order.To overcome the challenges, we first propose the Fact-aware Sentence Encoding, which enables the model to learn facts from the long sentence and thus improves the precision of sentence split; then we introduce Permutation Invariant Training to alleviate the effects of order variance in seq2seq learning for this task. Experiments on the WebSplit-v1.0 benchmark dataset show that our approaches can largely improve the performance over the previous seq2seq learning approaches. Moreover, an extrinsic evaluation on oie-benchmark verifies the effectiveness of our approaches by an observation that splitting long sentences with our state-of-the-art model as preprocessing is helpful for improving OpenIE performance.

MaskGEC: Improving Neural Grammatical Error Correction via Dynamic Masking

Grammatical error correction (GEC) is a promising natural language processing (NLP) application, whose goal is to change the sentences with grammatical errors into the correct ones. Neural machine translation (NMT) approaches have been widely applied to this translation-like task. However, such methods need a fairly large parallel corpus of error-annotated sentence pairs, which is not easy to get especially in the field of Chinese grammatical error correction. In this paper, we propose a simple yet effective method to improve the NMT-based GEC models by dynamic masking. By adding random masks to the original source sentences dynamically in the training procedure, more diverse instances of error-corrected sentence pairs are generated to enhance the generalization ability of the grammatical error correction model without additional data. The experiments on NLPCC 2018 Task 2 show that our MaskGEC model improves the performance of the neural GEC models. Besides, our single model for Chinese GEC outperforms the current state-of-the-art ensemble system in NLPCC 2018 Task 2 without any extra knowledge.

Simplify-Then-Translate: Automatic Preprocessing for Black-Box Translation

Black-box machine translation systems have proven incredibly useful for a variety of applications yet by design are hard to adapt, tune to a specific domain, or build on top of. In this work, we introduce a method to improve such systems via automatic pre-processing (APP) using sentence simplification. We first propose a method to automatically generate a large in-domain paraphrase corpus through back-translation with a black-box MT system, which is used to train a paraphrase model that “simplifies” the original sentence to be more conducive for translation. The model is used to preprocess source sentences of multiple low-resource language pairs. We show that this preprocessing leads to better translation performance as compared to non-preprocessed source sentences. We further perform side-by-side human evaluation to verify that translations of the simplified sentences are better than the original ones. Finally, we provide some guidance on recommended language pairs for generating the simplification model corpora by investigating the relationship between ease of translation of a language pair (as measured by BLEU) and quality of the resulting simplification model from back-translations of this language pair (as measured by SARI), and tie this into the downstream task of low-resource translation.

Lexical Simplification with Pretrained Encoders

Lexical simplification (LS) aims to replace complex words in a given sentence with their simpler alternatives of equivalent meaning. Recently unsupervised lexical simplification approaches only rely on the complex word itself regardless of the given sentence to generate candidate substitutions, which will inevitably produce a large number of spurious candidates. We present a simple LS approach that makes use of the Bidirectional Encoder Representations from Transformers (BERT) which can consider both the given sentence and the complex word during generating candidate substitutions for the complex word. Specifically, we mask the complex word of the original sentence for feeding into the BERT to predict the masked token. The predicted results will be used as candidate substitutions. Despite being entirely unsupervised, experimental results show that our approach obtains obvious improvement compared with these baselines leveraging linguistic databases and parallel corpus, outperforming the state-of-the-art by more than 12 Accuracy points on three well-known benchmarks.

Multi-Task Self-Supervised Learning for Disfluency Detection

Most existing approaches to disfluency detection heavily rely on human-annotated data, which is expensive to obtain in practice. To tackle the training data bottleneck, we investigate methods for combining multiple self-supervised tasks-i.e., supervised tasks where data can be collected without manual labeling. First, we construct large-scale pseudo training data by randomly adding or deleting words from unlabeled news data, and propose two self-supervised pre-training tasks: (i) tagging task to detect the added noisy words. (ii) sentence classification to distinguish original sentences from grammatically-incorrect sentences. We then combine these two tasks to jointly train a network. The pre-trained network is then fine-tuned using human-annotated disfluency detection training data. Experimental results on the commonly used English Switchboard test set show that our approach can achieve competitive performance compared to the previous systems (trained using the full dataset) by using less than 1% (1000 sentences) of the training data. Our method trained on the full dataset significantly outperforms previous methods, reducing the error by 21% on English Switchboard.

Grouping sentences as better language unit for extractive text summarization

Most existing methods for extractive text summarization aim to extract important sentences with statistical or linguistic techniques and concatenate these sentences as a summary. However, the extracted sentences are usually incoherent. The problem becomes worse when the source text and the summary are long and based on logical reasoning. The motivation of this paper is to answer the following two related questions: What is the best language unit for constructing a summary that is coherent and understandable? How is the extractive summarization process based on the language unit? Extracting larger language units such as a group of sentences or a paragraph is a natural way to improve the readability of summary as it is rational to assume that the original sentences within a larger language unit are coherent. This paper proposes a framework for group-based text summarization that clusters semantically related sentences into groups based on Semantic Link Network (SLN) and then ranks the groups and concatenates the top-ranked ones into a summary. A two-layer SLN model is used to generate and rank groups with semantic links including the is-part-of link, sequential link, similar-to link, and cause–effect link. The experimental results show that summaries composed by group or paragraph tend to contain more key words or phrases than summaries composed by sentences and summaries composed by groups contain more key words or phrases than those composed by paragraphs especially when the average length of source texts is from 7000 words to 17,000 words which is the usual length of scientific papers. Further, we compare seven clustering algorithms for generating groups and propose five strategies for generating groups with the four types of semantic links.

Personalized Sentence Generation using Generative Adversarial Networks with Author-Specific Word Usage

The author-specific word usage is a vital feature to let readers perceive the writing style of the author. In this work, a personalized sentence generation method based on generative adversarial networks (GANs) is proposed to cope with this issue. The frequently used function word and content word are incorporated not only as the input features but also as the sentence structure constraint for the GAN training. For the sentence generation with the related topics decided by the user, the Named Entity Recognition (NER) information of the input words is also used in the network training. We compared the proposed method with the GAN-based sentence generation methods, and the experimental results showed that the generated sentences using our method are more similar to the original sentences of the same author based on the objective evaluation such as BLEU and SimHash score.

Personalized Sentence Generation using Generative Adversarial Networks with Author-Specific Word Usage

The author-specific word usage is a vital feature to let readers perceive the writing style of the author. In this work, a personalized sentence generation method based on generative adversarial networks (GANs) is proposed to cope with this issue. The frequently used function word and content word are incorporated not only as the input features but also as the sentence structure constraint for the GAN training. For the sentence generation with the related topics decided by the user, the Named Entity Recognition (NER) information of the input words is also used in the network training. We compared the proposed method with the GAN-based sentence generation methods, and the experimental results showed that the generated sentences using our method are more similar to the original sentences of the same author based on the objective evaluation such as BLEU and SimHash score.

Development of the Everyday Conversational Sentences in Noise test.

To capture the demands of real-world listening, laboratory-based speech-in-noise tasks must better reflect the types of speech and environments listeners encounter in everyday life. This article reports the development of original sentence materials that were produced spontaneously with varying vocal efforts. These sentences were extracted from conversations between a talker pair (female/male) communicating in different realistic acoustic environments to elicit normal, raised and loud vocal efforts. In total, 384 sentences were extracted to provide four equivalent lists of 16 sentences at the three efforts for the two talkers. The sentences were presented to 32 young, normally hearing participants in stationary noise at five signal-to-noise ratios from -8 to 0 dB in 2 dB steps. Psychometric functions were fitted for each sentence, revealing an average 50% speech reception threshold (SRT50) of -5.2 dB, and an average slope of 17.2%/dB. Sentences were then level-normalised to adjust their individual SRT50 to the mean (-5.2 dB). The sentences may be combined with realistic background noise to provide an assessment method that better captures the perceptual demands of everyday communication.

A CNN-LSTM network with attention approach for learning universal sentence representation in embedded system

The model for obtaining universal sentence representation is getting larger and larger, making it unsuitable for small embedded systems. The paper presents an extended encoder-decoder model with introduced an attention mechanism for learning distributed sentence representation. We can extract the CNN encoder and apply to other NLP downstream tasks on small embedded systems. Inspired by the linguistic features of the word embeddings, the different dimensions of the sentence representation can be aligned to especially linguistic features. The decoder which decodes one word will focus on the partial dimension of sentence representation into a fixed-length vector where CNN is more effective than LSTM, especially on devices with limited computing power. Moreover, the expanded LSTM with attention mechanism as the decoder to learn multitask that reconstruct the original sentence and predict the next sentence. The model was trained on an extensive collection of a novel to learning sentence representation encoder. Finally, the small-scale CNN encoder obtained encouraging results on several benchmark datasets and multiple task.

Corrigendum to Fear of movement in patients attending cardiac rehabilitation: A validation study.

In the Abstract the following is sentence is changed: Original abstract sentence A strong negative correlation was found between the TSK-NL Heart and the HADS (Anxiety) (rs –0.51). Corrected abstract sentence Strong positive correlations were found between the TSK-NL Heart and de HADS (Anxiety) (rs: 0.60) and between the TSK-NL Heart and the CAQ (rs: 0.61).

СИНТАКСИЧНІ ТРАНСФОРМАЦІЇ У ПЕРЕКЛАДІ НАУКОВОЇ СТАТТІ

Publishing the research results in a science article with an international professional journal is an optimal way of sharing the information about newest discoveries in the world of science and technology. Not all scientists have a command of English sufficient for writing a science article, in compliance with high language requirements of leading scientific journals. So, the services of highly-qualified translators of scientific texts into English are in great request, and Ukraine is not an exception. Apart from the basic components of translator’s professional competence, especially important is the knowledge of norms of the modern English language scientific discourse. A translator of scientific texts is to have solid knowledge of grammar of source and target languages, regularities in rendering grammar forms and constructions, translation transformations. The largest number of grammar problems in translation is related to understanding the syntactic structure of sentences and a translator’s ability to make necessary transformations. Our research was aimed at analyzing the role of syntactic transformations in reaching the adequacy in English translation of Ukrainian language articles from scientometric journals. The analysis proved that the majority of syntactic transformations were used to bring the source text in conformity with the target language norms. The measure of translation transformations was generally adequate, though there were cases of non-use of syntactic transformations where they were necessary. Grammar literalism was also observed, due to translator’s insufficient understanding of the sentence structure, lack of knowledge of grammar peculiarities of the target language and translation solutions available for solving a particular translation problem. A translator of scientific texts should be particularly attentive to the syntax of the original sentence, analyse it properly, identify grammar phenomena that may cause translation problems and may need syntactic transformations, and build a translated sentence in accordance with the science language norms.

Explore and Practice of China's Intelligent “New Engineering” — Based on the Grounded Theory

In the era of the Industrial Revolution 4.0, emerging intelligent information technologies represented by the Internet of Things, big data, artificial intelligence, etc. are triggering a new round of educational reforms, driving human education to transition to intelligent education. This article adopts the 612 national-level new engineering research and practice projects released by China in 2018 that involve the integration of new engineering and wisdom education in universities and colleges that also implement the integration of the two in addition to the New Engineering Research and Practice Project，A total of 80 colleges and universities practice samples as research objects，With the help of the grounded theory system, a fusion model consisting of 755 original sentences, 77 concepts, 25categories, 7main categories, and 3 core categories-the TCG model. Its integration path is: Block chain: more open and more credible new ideas; Internet + education: the construction of a curriculum system that combines theory and practice with individuality and multiple coexistence; emotional skills perception + cloud computing: intelligent new teachers Strength training; AI + VR: construction of an open and immersive second learning world; big data + the Internet of Things: the establishment of a precise and intelligent management system; big data + artificial intelligence: the construction of a new mechanism of evaluation and incentives; the Internet of Things + none Seamless Interconnection: Comprehensive Perception of Government-Industry-Research Cooperation System.

The Involvement Load Hypothesis and Vocabulary Retention: The Interplay of Task Types and Involvement Index on L2 Vocabulary Learning

This study examines the claim that vocabulary learning and retention are dependent on a task's involvement load (i.e., need, search, evaluation), as proposed by Hulstijn and Laufer (2001a). The study aims at comparing the effects of task types and task involvement load on vocabulary retention for Korean EFL university students. More specifically, this study was designed to test whether differential levels of task involvement loads lead to equally effective results to vocabulary retention when the total involvement index being equal. Three types of productive word-focused tasks (gap-filling using a dictionary, writing original sentences, and gap-filling through word transformation) were used to examine the interplay of involvement index and task types. The result indicated that there were significant main effects of task types, test types, and proficiency levels. The results also indicated that there were significant interaction effects of task types on the retention tests, proficiency levels on the retention tests, and task types×proficiency levels on the retentions. The pedagogical implications and further research directions are discussed.

Expression of localization in the translations of Georg Wenker’s fourth sentence

This study examines the variation of the means of expressing localisation in different translations of one German sentence into Lithuanian. The main instrument of the analysis is one of the forty German sentences created in the 19th century by Georg Wenker, the pioneer of linguistic geography. The choice of this sentence was determined by its distinctiveness as it features specific characteristics of the Landmark and the interrelations between complex spatial scenes realised in it. The Lithuanian translations of the original German sentence were obtained by means of an electronic questionnaire. The goals of the study are to examine the inventory of the means of expressing localisation with respect to the parameter of the translator’s dialect and to verify the proposition made in prior studies of the Lithuanian translations of Wenker’s data that the variation of the means of expressing localisation is determined by the complex spatial scenes realised in the original sentence. A general evaluation of the newly acquired data and its analysis seems to confirm the proposition that the complex spatial scenes realised in the original sentence play a key role in determining the variation of the means of expressing localisation in the Lithuanian translations of Wenker’s fourth sentence. A comparison of the data acquired for the present study with the material obtained by Wenker in the 19th century leads to a conclusion that the linguistic inventory utilised to express spatial relations in the Lithuanian translations is essentially the same, i.e. Lithuanian prefers prepositional constructions as the prototypical means of rendering spatial relations. However, the newly obtained material contains more cases which indicate the possible influence of the translator’s dialect in the choice of the means of expressing localisation (cf. the use of the illative case, specific lexemes, etc.). In addition, the data collected for the present study also exhibits a greater variability of the means of expressing localisation used to refer specific spatial scenes. This could have been determined by the factor of subjectivity, namely, by the tendency of the respondents of this study to interpret the original sentence with greater freedom in comparison to its interpretation by the participants of Wenker’s study.

Computer Simulation of Surface‐Initiated Controlled Radical Polymerization: Effect of Free‐Monomer Model on Brush Properties

Macromol. Theory Simul. 2019, 28, 1900033 DOI: 10.1002/mats.201900033 In the original published article one sentence in the paper is inaccurate. The original sentence (page 2, right column, second paragraph, second sentence) states: “The authors observed that after polymerization stopped, the polymer chains relaxed their conformations and the concentration profiles broadened. That latter observation is the opposite of the ones we observed with a similar measurement described near the end of this manuscript.” The correct version of the sentence should read: “The authors observed that after polymerization stopped, the polymer chains relaxed their conformations and the concentration profiles broadened, in agreement with similar measurements described near the end of this manuscript.”

Sentence guided object color change by adversarial learning

In this paper, we propose a novel problem that is sentence guided object color change. Based on an original sentence and the corresponding image, by modifying this original sentence, the object color of corresponding image is changed along with the modified target sentence. This problem has two difficulties: (1) How to design a model to learn the change of sentence? (2) How to balance the relationship between the local object and the whole image during learning process? Confronted with these difficulties, as far as we know, few existing methods deal with them effectively. Therefore, we propose a new cascaded model to solve this problem based on generative adversarial networks (GANs). We employ the adversarial game to build a cascaded model, which learns the changed information of a sentence. Then, we specially design a penalty balance term to balance the relationship between local object and entire image in the generated image. Finally, experimental results on the flower and bird datasets demonstrate the validity of the proposed model.

An unsupervised constrained optimization approach to compressive summarization

Automatic summarization is typically aimed at selecting as much information as possible from text documents using a predefined number of words. Extracting complete sentences into a summary is not an optimal way to solve this problem due to redundant information that is contained in some sentences. Removing the redundant information and compiling a summary from compressed sentences should provide a much more accurate result. Major challenges of compressive approaches include the cost of creating large summarization corpora for training the supervised methods, the linguistic quality of compressed sentences, the coverage of the relevant content, and the time complexity of the compression procedure. In this work, we attempt to address these challenges by proposing an unsupervised polynomial-time compressive summarization algorithm. The proposed algorithm iteratively removes redundant parts from original sentences. It uses constituency-based parse trees and hand-crafted rules for generating elementary discourse units (EDUs) from their subtrees (standing for phrases) and selects ones with a sufficient tree gain. We define a parse tree gain as a weighted function of its node weights, which can be computed by any extractive summarization model capable of assigning importance weights to terms. The results of automatic evaluations on a single-document summarization task confirm that the proposed sentence compression procedure helps to avoid redundant information in the generated summaries. Furthermore, the results of human evaluations confirm that the linguistic quality—in terms of readability and coherency—is preserved in the compressed summaries while improving their coverage. However, the same evaluations show that compression in general harms the grammatical correctness of compressed sentences though, in most cases, this effect is not significant for the proposed compression procedure.