Sentence Segmentation Research Articles

Automatic Extraction of Flooding Control Knowledge from Rich Literature Texts Using Deep Learning

Flood control is a global problem; increasing number of flooding disasters occur annually induced by global climate change and extreme weather events. Flood studies are important knowledge sources for flood risk reduction and have been recorded in the academic literature. The main objective of this paper was to acquire flood control knowledge from long-tail data of the literature by using deep learning techniques. Screening was conducted to obtain 4742 flood-related academic documents from past two decades. Machine learning was conducted to parse the documents, and 347 sample data points from different years were collected for sentence segmentation (approximately 61,000 sentences) and manual annotation. Traditional machine learning (NB, LR, SVM, and RF) and artificial neural network-based deep learning algorithms (Bert, Bert-CNN, Bert-RNN, and ERNIE) were implemented for model training, and complete sentence-level knowledge extraction was conducted in batches. The results revealed that artificial neural network-based deep learning methods exhibit better performance than traditional machine learning methods in terms of accuracy, but their training time is much longer. Based on comprehensive feature extraction capability and computational efficiency, the performances of deep learning methods were ranked as: ERNIE > Bert-CNN > Bert > Bert-RNN. When using Bert as the benchmark model, several deformation models showed applicable characteristics. Bert, Bert-CNN, and Bert-RNN were good at acquiring global features, local features, and processing variable-length inputs, respectively. ERNIE showed improved masking mechanism and corpus and therefore exhibited better performance. Finally, 124,196 usage method and 8935 quotation method sentences were obtained in batches. The proportions of method sentence in the literature showed increasing trends over the last 20 years. Thus, as literature with more method sentences accumulates, this study lays a foundation for knowledge extraction in the future.

The Effect of Pre-segmented Instruction on Reading Fluency: A Study on Chinese-English Intermediate Learners

Intra–sentence segmentation has been proven a useful way for first language (L1) learners to increase reading fluency, including comprehension and reading rate. However, although it is widely used in China to deal with long sentences, it is still worth investigating whether it can be an effective strategy for Chinese-English intermediate learners. In addition, different methods used for segmenting also deserve to be studied. In this study, two English-speaking graduates and four Chinese- English learners are presented with nine long sentences within three different conditions (non-adjustable, segmenting different phrases with a backslash, and putting different phrases into different lines) and under each sentence there is a statement of correct or incorrect judgment. According to the results, the presegmented strategy increases the comprehension for Chinese- English intermediate learners and putting different phrases into different lines is most useful.

Product review opinion based on sentiment analysis

Sentiment analysis is a natural language processing (NLP) technique for determining emotional tone in a body of text. Using product reviews in sentiment analysis and opinion mining various methods have been developed previously. Although, existing product review analyzing techniques could not accurately detect the product aspect and non-aspect. Hence a novel Detach Frequency Assort is proposed to detect the product aspect term using TF-ISF (Term frequency-inverse sentence frequency) with Part of Speech (POS) tags for sentence segmentation and additionally using Feedback Neural Network to combine product aspect feedback loop. Furthermore, decision-making problem occurs during classification of sentiments. Hence, to solve this problem a novel technique named, Systemize Polarity Shift is proposed in which flow search based Support Vector Machine (SVM) with Bag of Words model classifies pre-trained review comments as positive, negative, and neutral sentiments. Moreover, the identification of specific products is not focused in sentiment analysis. Hence, a novel Revival Extraction is proposed in which a specific product is extracted based on thematic analysis method to obtain accurate data. Thus, the proposed Product Review Opinion framework gives effective optimized results in sentiment analysis with high accuracy, specificity, recall, sensitivity, F1-Score, and precision.

Using Natural Language Processing to Identify Low Back Pain in Imaging Reports

A natural language processing (NLP) pipeline was developed to identify lumbar spine imaging findings associated with low back pain (LBP) in X-radiation (X-ray), computed tomography (CT), and magnetic resonance imaging (MRI) reports. A total of 18,640 report datasets were randomly sampled (stratified by imaging modality) to obtain a balanced sample of 300 X-ray, 300 CT, and 300 MRI reports. A total of 23 radiologic findings potentially related to LBP were defined, and their presence was extracted from radiologic reports. In developing NLP pipelines, section and sentence segmentation from the radiology reports was performed using a rule-based method, including regular expression with negation detection. Datasets were randomly split into 80% for development and 20% for testing to evaluate the model’s extraction performance. The performance of the NLP pipeline was evaluated by using recall, precision, accuracy, and the F1 score. In evaluating NLP model performances, four parameters—recall, precision, accuracy, and F1 score—were greater than 0.9 for all 23 radiologic findings. These four scores were 1.0 for 10 radiologic findings (listhesis, annular fissure, disc bulge, disc extrusion, disc protrusion, endplate edema or Type 1 Modic change, lateral recess stenosis, Schmorl’s node, osteophyte, and any stenosis). In the seven potentially clinically important radiologic findings, the F1 score ranged from 0.9882 to 1.0. In this study, a rule-based NLP system identifying 23 findings related to LBP from X-ray, CT, and MRI reports was developed, and it presented good performance in regards to the four scoring parameters.

On the Performability of Drama Translation: A Case Study of Zhu Shenghao’s and Liang Shiqiu’s Versions of A Midsummer Night’s Dream

Drama translation, as a part of visual art, is more than just textual translation; it is a component of dramatic performance portrayed in the form of verbal action. Lines in drama translation are thus dialogues that render the drama performance rather than descriptive languages that provide information. Translators must use a variety of translation strategies to achieve performability in drama translation. Following Levý’s (2011) translation theory, the paper conducts a contrastive analysis of Zhu Shenghao’s and Liang Shiqiu’s Chinese versions of A Midsummer Night’s Dream from the aspects of the acquisition of speakability and intelligibility, restoration of semantic contexts, reflection of verbal action, and achievement of characterisation, exploring how the performability is satisfied in drama translation, and summarising strategies for fulfilling performability in drama translation. It is discovered that Zhu Shenghao (Zhu) insisted on his principles for drama translation, placed emphasis on performability in his translation, and used the following strategies to achieve performability in his translation: using colloquial style language, adjusting the order of sentence segments, reconstructing the syntax, uncovering the indeterminacy for different contexts, using modal particles, using dialects of the target language, and revealing characters’ intentions. However, in order to achieve speakability and intelligibility, Zhu abandoned the stichomythia rhyme, failing to present the verbal action. On the other hand, Liang preserved the stichomythia rhyme, reflecting the verbal action. Furthermore, Zhu replaced the personal pronoun in the original text with notional terms, resulting in the loss of semantic contexts, whereas Liang Shiqiu (Liang) preserved the personal pronoun, restoring the semantic contexts.

English Long Sentence Segmentation and Translation Optimization of Professional Literature Based on Hierarchical Network of Concepts

The structure of English long sentences is generally complicated, with complicated logical levels, many parallel elements, many modifiers and conjunctions, and long post attributives. Moreover, pronouns in English long sentences often need to be judged according to the context. Complex English long sentences are very common in English and even run through the whole English article. The analysis results of these complex long sentences will seriously affect the quality and readability of machine translation. In this paper, HNC (hierarchical network of concepts) method is improved to realize the segmentation of long sentences, so as to simplify sentence patterns. According to the characteristics of professional literature, this paper puts forward a translation optimization method of professional literature combining HNC with statistics, which greatly improves the recognition efficiency of unknown words by extracting professional terms. The results show that the word segmentation method based on HNC and statistics proposed in this paper has achieved a good word segmentation effect in the open test environment, with an accuracy rate of 93.38% and a recall rate of 94.51%. The conclusion shows that our method can make full use of the knowledge of the source tree database, thereby improving the accuracy of the syntactic model on the target tree database.

Aspect Based Opinion Summarization Using Rule-Based Method and Support Vector Machine for Indonesian Reviews

E-commerce's reviews feature can help users find information about the desired skin care products to choose the right one. However, the reviews number of a product grows rapidly due to the popularity of e-commerce and the product itself. A user becomes difficult to read all reviews one by one and extract useful information. To deal with this problem, we summarize aspects using the Rule-Based method and Support Vector Machine. We propose a Rule-Based method that is used to break down a review into several segments based on its aspect. Support Vector Machine is used to classify sentence segments according to their polarity. The data used in this study is Indonesian reviews of skin care products obtained from the Female Daily website. The average accuracy results using 10-fold cross-validation of sentiment classification is 74%. We experimented on 462 reviews where the accuracy is 92% in aspect categorization and 71.2% in sentiment classification. Based on humans, the lowest value is the suitability of a sentence with its sentiment/polarity. The highest value is the suitability of sentences with its aspect and usability of summary to helps users to find specific information so they can decide whether to buy that product or not. It can be concluded that the reader can well receive the summary. Future work can consider the negation word to reduce misclassification in the sentiment classification step.

E-Commerce Brand Ranking Algorithm Based on User Evaluation and Sentiment Analysis.

ObjectiveConsumers often need to compare the same type of products from different merchants to determine their purchasing needs. Fully mining the product information on the website and applying it to e-commerce websites or product introduction websites can not only allow consumers to buy products that are more in line with their wishes, but also help merchants understand user needs and the advantages of each product. How to quantify the emotional tendency of evaluation information and how to recommend satisfactory products to consumers is the research purpose of this paper.MethodAccording to the analysis of the research object, this paper uses the Python crawler library to efficiently crawl the data required for research. By writing a custom program for crawling, the resulting data is more in line with the actual situation. This paper uses the BeautifulSoup library in Python web crawler technology for data acquisition. Then, in order to ensure high-quality data sets, the acquired data needs to be cleaned and deduplicated. Finally, preprocessing such as sentence segmentation, word segmentation, and semantic analysis is performed on the cleaned data, and the data format required by the subsequent model is output. For weightless network, the concept of node similarity is proposed, which is used to measure the degree of mutual influence between nodes. Combined with the LeaderRank algorithm, and fully considering the differences between nodes in the interaction, the SRank algorithm is proposed. Different from the classical node importance ranking method, the SRank algorithm fully considers the local and global characteristics of nodes, which is more in line with the actual network.Results/DiscussionThis paper calculates the sentiment polarity of users’ comments, obtains the final user influence ranking, and identifies opinion leaders. The final ranking results were compared and analyzed with the traditional PageRank algorithm and SRank ranking algorithm, and it was found that the opinion leaders identified by the opinion leader identification model integrating user activity and comment sentiment were more reasonable and objective. The algorithm in this paper improves the efficiency of operation to a certain extent, and at the same time solves the problem that sentiment analysis cannot be effectively used in social network analysis, and increases the accuracy of e-commerce brand ranking.

Abstract Translation in Scientific Writing by Non-English Major Student using Google Translate

The main purpose of this research is to find out Google Translate errors when translating bachelor's paper abstracts from Indonesian to English. Six abstracts of Non-English major undergraduate student papers from diverse faculties Sekolah Tinggi Ilmu Kesehatan Siti Khadijah Palembang were chosen at random. Following that, the data is compared for each sentence segment, as well as any words or phrases with problems are investigated. The result shows that the primary errors of abstract translations include lexicosemantic, tense, preposition, word order, distribution, and use of verb group, as well as active and passive voice errors. Abstract translation outcomes deviate significantly from expectations. Both translating English writings into Indonesian and Indonesian texts into English takes advanced translation abilities. Regarding translation, several factors should be considered; they include the linguistic system utilized for both languages and the cultural understanding of the two languages. In the case of scientific writing translation, involve a translator in improving the quality of translation. In terms of academics, it is recommended that further research be done on comparable challenges in abstract translation, but with more variety from diverse sources.

Natural language generation and deep learning for intelligent building codes

Many existing automated compliance checking (ACC) systems require the processes of extracting regulatory information from natural-language building-code requirements and transforming the extracted information into computer-processable semantic representations. These processes could, however, be jeopardized by the ambiguous nature of the natural language and the hierarchically complex structures of building-code requirements. To address this problem, this paper proposes the concept of intelligent building code for bypassing the error-prone information extraction and transformation processes. In the proposed intelligent code, the natural-language requirements in the code are connected with highly structured computer-understandable semantic information, which is represented in the form of semantic requirement hierarchies and can be readily used by computers for ACC. The paper also proposes a deep learning-based method to automatically generate such intelligent code. The method leverages the requirement hierarchy representation, a proposed deep learning unit-to-text model for generating requirement sentence segments, and a proposed semantic correspondence score for configuring the segments into requirement sentences. The method was implemented and tested on a dataset from multiple regulatory documents. The generated intelligent requirements were evaluated in terms of both natural-language requirement comprehensibility and correspondence between the natural language and the semantic representation, with the results indicating high performance for the proposed representation and method. The proposed intelligent code will help reduce ACC errors, improve requirement comprehensibility, and facilitate intelligent code analytics.

APLIKASI PERINGKASAN DOKUMEN MENGGUNAKAN METODE MAXIMUM MARGINAL RELEVANCE (MMR)

making a summary application to help readers who do not like to read long and thick news articles which take a relatively long time and can cause readers to be lazy to read the news articles. This summary application is used to summarize news articles, in making the application using ASP.Net. In this summary, the Maximum Marginal Relevance (MMR) method is used. In this study, you can use articles on the website, and do it. Articles are processed in the form of a file (single document) with a txt extension. The summary process goes through the preprocessing stage, which consists of sentence segmentation, case folding, tokenizing, filtering, stemming.

Типологічна характеристика логічної сегментації речень української та англійської мови

The proposed article is devoted to the comparative study of the means of expression of the logical segmentation in English and Ukrainian. The topic of the article is determined by the fact that the study of functional perspective of a sentence is now the most important stage in the formation of communicative and functional grammar of any language. The article analyzes the general theoretic foundations of the notion of “segmentation”. The main features of sentence segmentation in linguistics are characterized. In the course of the work main stages of scientific thought in the field of comparative linguistics are reflected, such concepts as “segmentation”, “logical segmentation” are studied, and principles of the logical segmentation of sentences in the languages under consideration are analysed. A systematic analysis of the available data on the problem of segmentation is performed. The means of segmentation expression in the languages are identified and analyzed. It has been revealed that the complex of the main differences in the segmentation of sentences in Ukrainian and English languages. The complex analysis of pedagogical, psychological and linguistic knowledge of the research problem has been established. The mechanism of segmentation, which is one of the special mechanisms providing all aspects of communication, has been revealed. Real links in a sentence are analysed and described. The knowledge on the peculiarities of segmentation of the English-language text has been identified, summarized and systematized. A comparative-typological description of the two languages on the problem has been carried out. Based on these theories, the main problems of sentence segmentation in English are highlighted. Based on the comparison with the English language, the syntactic, grammatical and lexical ways of separating the topic and the rheme are highlighted and described in detail in a comparative perspective. In the course of the study, the universal mechanisms of sentence segmentation in English and Ukrainian are revealed, as well as the areas that are selective in either English or Ukrainian. The article considers the phenomenon of segmentation in different aspects: the history of the development of structures as expressive units of literary language, the types of segmentation are analyzed.

Similitude Based Segment Graph Construction and Segment Ranking for Automatic Summarization of Text Document

With the increase in the amount of data and documents on the web, text summarization has become one of the significant fields which cannot be avoided in today’s digital era. Automatic text summarization provides a quick summary to the user based on the information presented in the text documents. This paper presents the automated single document summarization by constructing similitude graphs from the extracted text segments. On extracting the text segments, the feature values are computed for all the segments by comparing them with the title and the entire document and by computing segment significance using the information gain ratio. Based on the computed features, the similarity between the segments is evaluated to construct the graph in which the vertices are the segments and the edges specify the similarity between them. The segments are ranked for including them in the extractive summary by computing the graph score and the sentence segment score. The experimental analysis has been performed using ROUGE metrics and the results are analyzed for the proposed model. The proposed model has been compared with the various existing models using 4 different datasets in which the proposed model acquired top 2 positions with the average rank computed on various metrics such as precision, recall, F-score.
 HIGHLIGHTS
 
 Paper presents the automated single document summarization by constructing similitude graphs from the extracted text segments
 It utilizes information gain ratio, graph construction, graph score and the sentence segment score computation
 Results analysis has been performed using ROUGE metrics with 4 popular datasets in the document summarization domain
 The model acquired top 2 positions with the average rank computed on various metrics such as precision, recall, F-score
 
 GRAPHICAL ABSTRACT

METHOD OF DOMAIN ONTOLOGY AUTOMATED REPLENISHMENT FOR THE SUPPORT OF NEW TECHNICAL SOLUTIONS SYNTHESIS. PART I

To solve the problem of information support for the synthesis of new technical solutions, a method of extracting structured data from an array of Russian-language patents is presented. The key features of the invention, such as the structural elements of the technical object and the relationships between them, are considered as information support. The data source addresses the main claim of the invention in the device patent. The unit of extraction is the semantic structure Subject-Action-Object (SAO), which semantically describes the constructive elements. The extraction method is based on shallow parsing and claim segmentation, taking into account the specifics of writing patent texts. Often the excessive length of the claim sentence and the specificity of the patent language make it difficult to efficiently use off-the-shelf tools for data extracting. All processing steps include: segmentation of the claim sentences; extraction of primary SAO structures; construction of the graph of the construct elements f the invention; integration of the data into the domain ontology. This article deals with the first two stages. Segmentation is carried out according to a number of heuristic rules, and several natural language processing tools are used to reduce analysis errors. The primary SAO elements are extracted considering the valences of the predefined semantic group of verbs, as well as information about the type of processed segment. The result of the work is the organization of the domain ontology, which can be used to find alternative designs for nodes in a technical object. In the second part of the article, an algorithm for constructing a graph of structural elements of a separate technical object, an assessment of the effectiveness of the system, as well as ontology organization and the result are considered.

Feature Fusion Strategies for End-to-End Evaluation of Cognitive Behavior Therapy Sessions.

Cognitive Behavioral Therapy (CBT) is a goal-oriented psychotherapy for mental health concerns implemented in a conversational setting. The quality of a CBT session is typically assessed by trained human raters who manually assign pre-defined session-level behavioral codes. In this paper, we develop an end-to-end pipeline that converts speech audio to diarized and transcribed text and extracts linguistic features to code the CBT sessions automatically. We investigate both word-level and utterance-level features and propose feature fusion strategies to combine them. The utterance level features include dialog act tags as well as behavioral codes drawn from another well-known talk psychotherapy called Motivational Interviewing (MI). We propose a novel method to augment the word-based features with the utterance level tags for subsequent CBT code estimation. Experiments show that our new fusion strategy outperforms all the studied features, both when used individually and when fused by direct concatenation. We also find that incorporating a sentence segmentation module can further improve the overall system given the preponderance of multi-utterance conversational turns in CBT sessions.

Semi-supervised Thai Sentence Segmentation Using Local and Distant Word Representations

A sentence is typically treated as the minimal syntactic unit used for extracting valuable information from a longer piece of text. However, in written Thai, there are no explicit sentence markers. We proposed a deep learning model for the task of sentence segmentation that includes three main contributions. First, we integrate n-gram embedding as a local representation to capture word groups near sentence boundaries. Second, to focus on the keywords of dependent clauses, we combine the model with a distant representation obtained from self-attention modules. Finally, due to the scarcity of labeled data, for which annotation is difficult and time-consuming, we also investigate and adapt Cross-View Training (CVT) as a semi-supervised learning technique, allowing us to utilize unlabeled data to improve the model representations. In the Thai sentence segmentation experiments, our model reduced the relative error by 7.4% and 10.5% compared with the baseline models on the Orchid and UGWC datasets, respectively. We also applied our model to the task of pronunciation recovery on the IWSLT English dataset. Our model outperformed the prior sequence tagging models, achieving a relative error reduction of 2.5%. Ablation studies revealed that utilizing n-gram presentations was the main contributing factor for Thai, while the semi-supervised training helped the most for English.

Deep Learning Based Automatic Sentence Segmentation and Punctuation Model for Massive Classical Chinese Literature

Deep Learning Based Automatic Sentence Segmentation and Punctuation Model for Massive Classical Chinese Literature

Segmented Translation Algorithm of Complex Long Sentences Based on Semantic Features

The progress of the times is inseparable from communication. If a country wants to develop well, it must learn from each other. Accurate language translation can better let people understand what they want to express. Therefore, language translation is becoming more and more important in the current social communication. Although there is a lot of research on translation, in many cases there will be inaccurate translations. Therefore, finding an accurate translation method is what many people need. Aiming at the problem that the translation of complex long sentences is prone to errors, this paper proposes a sentence segmentation algorithm, which is a method of dividing the long sentence into multiple independent clauses and then translating it. The segmentation algorithm uses the semantic features of the Concept Hierarchical Network (HNC) theory to segment clauses. The segmentation algorithm is integrated with a rule-based baseline translation system. The BLEU value of the integrated translation system reaches 0.1898, which is higher than that before the integration. The system has increased by 30%. Experimental results prove that the proposed method can effectively improve the effect of patent translation.

APLIKASI CLUSTERING BERITA DENGAN METODE K MEANS DAN PERINGKAS BERITA DENGAN METODE MAXIMUM MARGINAL RELEVANCE

News is information about facts or opinions that are interesting to know. News can be obtained from various media such as newspapers and the internet. As is well known, news has various topics, such as politics, sports and others. There is also the same story written with the addition of a little information. This causes it to take more time to get the headline of the news. Therefore we need a system for news clustering using the K-Means method and news summarizing using the Maximum Marginal Relevance (MMR) method in order to obtain information from news more easily and efficiently. News that is processed in the form of a collection of files (multi document) with the extension txt. The summarization process goes through the text preprocessing stage, which consists of sentence segmentation, case folding, tokenizing, filtering, stemming. The next step is TF-IDF calculation to calculate word weight then Cosine Similarity to calculate the similarity between documents. After that, enter the K-Means stage for clustering division and proceed with determining the summary with MMR. Based on the results testing that has been done, this application is running well, the results of clustering and summarizing news can make it easier for users to get news summaries from some similar news.

AUTOMATED SYSTEM FOR DETECTION OF NON-STANDARD ACTIONS USING SCRIPTURAL ANALYSIS OF THE TEXT

The scenario (narrative schemas) is some established (in society) sequence of steps to achieve the set goal and contains the most complete information about all possible ways of development of the described situation (with selection points and branches). The creation of the XML platform gave rise to a new high-tech and technologically more advanced stage in the development of the Web. As a result, the XML platform becomes a significant component in the technology of information systems development, and the tendency of their integration at the level of corporations, agencies, ministries only strengthens the position of XML in the field of information technology in general. A system for automatic detection of non-standard scripts in text messages has been developed. System programming consists of stages of ontology formation, sentence parsing and scenario comparison. the classic natural language processing (NLP) method, which supports the most common tasks such as tokenization, sentence segmentation, tagging of a part of speech, extraction of named entities, partitioning, parsing and co-referential resolution, is used for parsing sentences in the system. Maximum entropy and machine learning based on perceptrons are also possible. Ontologies are stored using OWL technology. The object-target sentence parsers with the described OWL are compared in the analysis process. From a SPARQL query on a source object, query models are returned to the table object. The table class is the base class for all table objects and provides an interface for accessing values in the rows and columns of the results table. If the table object has exactly three columns, it can be used to build a new data source object. This provides a convenient mechanism for retrieving a subset of data from one data source and adding them to another. In the context of the RDF API, a node is defined as all statements about the subject of a URI. The content of the table is compared with the semantics of the sentence. If the sentence scenario does not match the OWL ontology model, there is a possibility of atypical object actions. In this case, a conclusion is formed about the suspicion of the message. For more correct use of possibilities of the analysis of the text it is necessary to form the case of ontologies or to use existing (Akutan, Amazon, etc.) taking into account their features. To increase the ontologies of objects, it is possible to use additional neural network teaching methods.