基于Sentence-BERT与孤立森林算法的专利新颖性评估

进行关键技术抽取作为实施方法特征;其次,采用Sentence-BERT对上述特征进行文本向量化表示后组合输入至孤立森林算法获得离群专利集;最后,通过技术量权值过滤法提高专利新颖性评估的精度。以金融科技领域专利进行实证研究,结果表明,该评估方法准确率相较专业专利分析平台方法提升了9%~11%。证明了该方法在专利新颖性评估中的有效性,能为后续专利审核工作和高价值专利分析提供参考。

Sentence及其结构研究——以古典主义时期作品为例

文章以sentence为研究对象,一方面对国内外曲式学教材及相关文献展开概念比较、辨析与理论梳理,阐明其基本概念与内部结构;另一方面则以古典主义时期作品为分析实例,讨论这种特定主题(句法)的“结构范型”、并进一步对其“结构变形”的情况进行归类。

Product image sentence annotation based on kernel descriptors and tag-rank

Dealing with issues such as too simple image features and word noise inference in product image sentence anmotation, a product image sentence annotation model focusing on image feature learning and key words summarization is described. Three kernel descriptors such as gradient, shape, and color are extracted, respectively. Feature late-fusion is executed in turn by the multiple kernel learning model to obtain more discriminant image features. Absolute rank and relative rank of the tag-rank model are used to boost the key words＇ weights. A new word integration algorithm named word sequence blocks building （WSBB） is designed to create N-gram word sequences. Sentences are generated according to the N-gram word sequences and predefined templates. Experimental results show that both the BLEU-1 scores and BLEU-2 scores of the sentences are superior to those of the state-of-art baselines.

基于Sentence-BERT的专利技术主题聚类研究——以人工智能领域为例

[研究目的]将Sentence-BERT模型应用于专利技术主题聚类,解决专利文献为突出新颖性,常使用独特技术术语造成词汇向量语义特征稀疏的问题。[研究方法]以人工智能领域2015年-2019年的22370篇专利为实验数据。首先,采用Sentence-BERT算法对专利文献摘要文本进行向量化表示;其次,对向量化矩阵进行数据降维,利用HDBSCAN方式寻找原始数据中的高密度簇;最后,识别类簇文本集合中的主题特征,并完成主题呈现。[研究结论]对比LDA主题模型、K-means、doc2vec等方法,本文的实验结果提高了主题划分的细粒度和精确度,获得了较好的主题一致性。如何采用fine-tune策略进一步提升模型的效果,是未来该方法进一步深入探索的方向。

Masked Sentence Model Based on BERT for Move Recognition in Medical Scientific Abstracts

Purpose:Mo ve recognition in scientific abstracts is an NLP task of classifying sentences of the abstracts into different types of language units.To improve the performance of move recognition in scientific abstracts,a novel model of move recognition is proposed that outperforms the BERT-based method.Design/methodology/approach:Prevalent models based on BERT for sentence classification often classify sentences without considering the context of the sentences.In this paper,inspired by the BERT masked language model(MLM),we propose a novel model called the masked sentence model that integrates the content and contextual information of the sentences in move recognition.Experiments are conducted on the benchmark dataset PubMed 20K RCT in three steps.Then,we compare our model with HSLN-RNN,BERT-based and SciBERT using the same dataset.Findings:Compared with the BERT-based and SciBERT models,the F1 score of our model outperforms them by 4.96%and 4.34%,respectively,which shows the feasibility and effectiveness of the novel model and the result of our model comes closest to the state-of-theart results of HSLN-RNN at present.Research limitations:The sequential features of move labels are not considered,which might be one of the reasons why HSLN-RNN has better performance.Our model is restricted to dealing with biomedical English literature because we use a dataset from PubMed,which is a typical biomedical database,to fine-tune our model.Practical implications:The proposed model is better and simpler in identifying move structures in scientific abstracts and is worthy of text classification experiments for capturing contextual features of sentences.Originality/value:T he study proposes a masked sentence model based on BERT that considers the contextual features of the sentences in abstracts in a new way.The performance of this classification model is significantly improved by rebuilding the input layer without changing the structure of neural networks.

CiteOpinion: Evidence-based Evaluation Tool for Academic Contributions of Research Papers Based on Citing Sentences

Purpose:To uncover the evaluation information on the academic contribution of research papers cited by peers based on the content cited by citing papers,and to provide an evidencebased tool for evaluating the academic value of cited papers.Design/methodology/approach:CiteOpinion uses a deep learning model to automatically extract citing sentences from representative citing papers;it starts with an analysis on the citing sentences,then it identifies major academic contribution points of the cited paper,positive/negative evaluations from citing authors and the changes in the subjects of subsequent citing authors by means of Recognizing Categories of Moves(problems,methods,conclusions,etc.),and sentiment analysis and topic clustering.Findings:Citing sentences in a citing paper contain substantial evidences useful for academic evaluation.They can also be used to objectively and authentically reveal the nature and degree of contribution of the cited paper reflected by citation,beyond simple citation statistics.Practical implications:The evidence-based evaluation tool CiteOpinion can provide an objective and in-depth academic value evaluation basis for the representative papers of scientific researchers,research teams,and institutions.Originality/value:No other similar practical tool is found in papers retrieved.Research limitations:There are difficulties in acquiring full text of citing papers.There is a need to refine the calculation based on the sentiment scores of citing sentences.Currently,the tool is only used for academic contribution evaluation,while its value in policy studies,technical application,and promotion of science is not yet tested.

A Sentence Similarity Estimation Method Based on Improved Siamese Network

In this paper we employ an improved Siamese neural network to assess the semantic similarity between sentences. Our model implements the function of inputting two sentences to obtain the similarity score. We design our model based on the Siamese network using deep Long Short-Term Memory (LSTM) Network. And we add the special attention mechanism to let the model give different words different attention while modeling sentences. The fully-connected layer is proposed to measure the complex sentence representations. Our results show that the accuracy is better than the baseline in 2016. Furthermore, it is showed that the model has the ability to model the sequence order, distribute reasonable attention and extract meanings of a sentence in different dimensions.

A Research on Bilingual Dictionary Based Sentence Alignment for Chinese English Parallel Corpus

Parallel corpus is of great importance to machine translation, and automatic sentence alignment is the first step towards its processing. This paper puts forward a bilingual dictionary based sentence alignment method for Chinese English parallel corpus, which differs from previous length based algorithm in its knowledge-rich approach. Experimental result shows that this method produces over 93% accuracy with usual English-Chinese dictionaries whose translations cover 31 88%～47 90% of the corpus.

English Sentence Recognition Based on HMM and Clustering

For English sentences with a large amount of feature data and complex pronunciation changes contrast to words, there are more problems existing in Hidden Markov Model (HMM), such as the computational complexity of the Viterbi algorithm and mixed Gaussian distribution probability. This article explores the segment-mean algorithm for dimensionality reduction of speech feature parameters, the clustering cross-grouping algorithm and the HMM grouping algorithm, which are proposed for the implementation of the speaker-independent English sentence recognition system based on HMM and clustering. The experimental result shows that, compared with the single HMM, it improves not only the recognition rate but also the recognition speed of the system.

Review of Research on English Translation of Chinese Running Sentences

In order to convey complete meanings,there is a phenomenon in Chinese of using multiple running sentences.Xu Jingning(2023,p.66)states,“In communication,a complete expression of meaning often requires more than one clause,which is common in human languages.”Domestic research on running sentences includes discussions on defining the concept and structural features of running sentences,sentence properties,sentence pattern classifications and their criteria,as well as issues related to translating running sentences into English.This article primarily focuses on scholarly research into the English translation of running sentences in China,highlighting recent achievements and identifying existing issues in the study of running sentence translation.However,by reviewing literature on the translation of running sentences,it is found that current research in the academic community on non-core running sentences is limited.Therefore,this paper proposes relevant strategies to address this issue.

Windowing and Gapping in Imperative Sentences: on the Basis of Talmy's"Causal-chain Windowing"Approach

This paper intends to analyze the six types of English imperative sentences proposed by Chen (1984) from a perspective of causal-chain windowing. It comes to the conclusions that Talmy's causal-chain windowing approach as well as the cognitive underpinnings of causal windowing and gapping is proved to be applicable in English imperative structures, and that generally speaking, the final portion of an imperative sentence is always windowed while the intermediate portions gapped.

我国民间体育赛事短视频传播效益形成机理及提升路径——以“村BA”为例

运用文献资料、案例分析、逻辑分析等方法,以“村BA”为例分析我国民间体育赛事短视频传播效益、形成机理,并提出提升路径。我国民间体育赛事短视频传播具有产生经济发展、生发文化融合、形成人际联结、孕育全新生态四方面的效益。形成机理:数字技术革新,拓宽传播流量;新老文化交融,创新内容形式;多元交互联结,增强用户黏性;传播门槛降低,构建三维生态。提升路径:挖掘音画素材内容,优化视频制作技术;融入赛事视觉系统,显现物质文化;强化联合监督管理,加大政策扶持;完善运营管理体系,打造全媒体矩阵。

Classification of Conversational Sentences Using an Ensemble Pre-Trained Language Model with the Fine-Tuned Parameter

Sentence classification is the process of categorizing a sentence based on the context of the sentence.Sentence categorization requires more semantic highlights than other tasks,such as dependence parsing,which requires more syntactic elements.Most existing strategies focus on the general semantics of a conversation without involving the context of the sentence,recognizing the progress and comparing impacts.An ensemble pre-trained language model was taken up here to classify the conversation sentences from the conversation corpus.The conversational sentences are classified into four categories:information,question,directive,and commission.These classification label sequences are for analyzing the conversation progress and predicting the pecking order of the conversation.Ensemble of Bidirectional Encoder for Representation of Transformer(BERT),Robustly Optimized BERT pretraining Approach(RoBERTa),Generative Pre-Trained Transformer(GPT),DistilBERT and Generalized Autoregressive Pretraining for Language Understanding(XLNet)models are trained on conversation corpus with hyperparameters.Hyperparameter tuning approach is carried out for better performance on sentence classification.This Ensemble of Pre-trained Language Models with a Hyperparameter Tuning(EPLM-HT)system is trained on an annotated conversation dataset.The proposed approach outperformed compared to the base BERT,GPT,DistilBERT and XLNet transformer models.The proposed ensemble model with the fine-tuned parameters achieved an F1_score of 0.88.

Selective Sentence Production Deficit in an Agrammatic Yoruba-English Bilingual Patient with Minor Stroke: A Case Study

Speech helps us to communicate with our loved ones and significant others through construction of grammatically coherent sentences that are comprehensible to our communication partners. As such, impairment of this ability as a result of stroke can be debilitating and disabling to the patients as well as significant others. Agrammatism is deficit in the use and processing of grammatically coherent syntactic structures following damage to the Broca’s complex or region. Most studies have traditionally emphasized monolingual patients, with bilingualism now receiving increased attention. However, few studies have specifically investigated the effect of minor stroke on agrammatic bilingual individuals. This study examined an agrammatic Yoruba- English bilingual patient with minor stroke with a view to describing their sentence production (deficit). The findings strongly support the existence of distinct language-specific lexical-subsystem centres in the Broca’s complex for native and acquired languages (Yoruba-English) whereas both languages are likely connected to a single semantic system in the anterior temporal lobe and its surrounding regions. Furthermore, acquired language is more susceptible to brain damage than native language. This might imply that severity of deficit in speech production in both native and acquired language of bilingual aphasics may be determined by the size of lesion in the Broca’s complex or region.

An Analysis Into Ba Sentences Under Image Schema Theory——A Case Study of Words of Fire——Poems by Jidi Majia

The Ba sentence is unique existence in Chinese. There is no corresponding sentence pattern in English. Thus, the translation of Ba sentences is a challenge for translators. Even with the myriad analyses on the translation of Ba sentences, small quantity of them demonstrates the correctness and the reading effect of the translation by using a specific theory. And the image schema theory reflects the projects between the source domain and the target domain, while in the translation analysis there are schemata in the source language and the target language. So the paper does comparisons of schemata between the source text and the target text of Ba sentences, which are chosen from the English translation of Words of Fire—Poems by Jidi Majia translated by Denis Mair. After the demonstration, the following conclusions are found: First, the schema of the sentence is decided by verbs of the Ba sentences, rather than by the sentence structures;second, the image schema is a feasible tool to check correctness of Ba sentences translation.

新发展理念下贵州省“村BA”特色篮球高质量发展路径

本文运用文献资料法、系统分析法等方法,深入研判贵州省“村BA”特色篮球发展的现状与困境,明晰新发展理念下该地区特色篮球发展的价值导向与优化路径。当前贵州省村BA特色篮球发展存在结构性治理矛盾复杂、政策性引导力度不足以及产业性品牌发展滞后的现实困境,将其与新发展理念中“创新、绿色、开放、共享”四大方面结合治理有利于形成助力乡村振兴、驱动改革创新及拓宽发展需求的有效路径。新发展理念下贵州省“村BA”特色篮球的高质量发展路径包括加强公共服务建设、优化人才培养体系、完善赛事治理模式、深挖品牌文化价值以及助力产业顶层设计五个维度,这是实现其高质量发展的保障。

Next Words Prediction and Sentence Completion in Bangla Language Using GRU-Based RNN on N-Gram Language Model

We use a lot of devices in our daily life to communicate with others. In this modern world, people use email, Facebook, Twitter, and many other social network sites for exchanging information. People lose their valuable time misspelling and retyping, and some people are not happy to type large sentences because they face unnecessary words or grammatical issues. So, for this reason, word predictive systems help to exchange textual information more quickly, easier, and comfortably for all people. These systems predict the next most probable words and give users to choose of the needed word from these suggested words. Word prediction can help the writer by predicting the next word and helping complete the sentence correctly. This research aims to forecast the most suitable next word to complete a sentence for any given context. In this research, we have worked on the Bangla language. We have presented a process that can expect the next maximum probable and proper words and suggest a complete sentence using predicted words. In this research, GRU-based RNN has been used on the N-gram dataset to develop the proposed model. We collected a large dataset using multiple sources in the Bangla language and also compared it to the other approaches that have been used such as LSTM, and Naive Bayes. But this suggested approach provides excellent exactness than others. Here, the Unigram model provides 88.22%, Bi-gram model is 99.24%, Tri-gram model is 97.69%, and 4-gram and 5-gram models provide 99.43% and 99.78% on average accurateness. We think that our proposed method profound impression on Bangla search engines.

On Translation of Long Sentences Based on the Novel-Of Time and the River

Of Time and the River is a famous novel written by American novelist–Thomas Wolfe who was a major American novelist of the early 20th century.His books are well received among Chinese readers.In the novel—Of Time and the River,we can find Wolfe’s frequent use of long sentences.In this paper,the skills of translating long sentences and the ways of analyzing sentence structures are discussed by several examples selected from the novel.Based on my own translation,we can find the importance of sentence structure analysis in translating long sentences.

Mother Tongue Interference in Japanese Passive Sentence Teaching —— Analysis Based on Functional Grammar

For foreign language learners, whether Chinese people are learning Japanese or Japanese learning Chinese, if they can eliminate the interference of basic knowledge of the mother tongue, the efficiency of foreign language learning will be greatly improved. This paper mainly analyzes the problem of mother tongue interference encountered by Japanese learners who use Chinese as their mother tongue in the study of "passive sentences". It focuses on introducing Japanese functional grammar on the basis of traditional grammar to eliminate these interference problems.

Research of Paraphrasing for Chinese Complex Sentences Based on Templates

Based on the paraphrasing of Chinese simple sentences,the complex sen­tence paraphrasing by using templates are studied.Through the classifica­tion of complex sentences,syntactic analysis and structural analysis,the proposed methods construct complex sentence paraphrasing templates that the associated words are as the core.The part of speech tagging is used in the calculation of the similarity between the paraphrasing sentences and the paraphrasing template.The joint complex sentence can be divided into parallel relationship,sequence relationship,selection relationship,progres­sive relationship,and interpretive relationship’s complex sentences.The subordinate complex sentence can be divided into transition relationship,conditional relationship,hypothesis relationship,causal relationship and objective relationship’s complex sentences.Joint complex sentence and subordinate complex sentence are divided to associated words.By using pretreated sentences,the preliminary experiment is carried out to decide the threshold between the paraphrasing sentence and the template.A small scale paraphrase experiment shows the method is availability,acquire the coverage rate of paraphrasing template 40.20%and the paraphrase correct rate 62.61%.