Word-level Analysis Research Articles

Lead News Coverage of Buddhist Monks in The Nation Online Newspaper: A Syntactic Analysis

This research conducts Lead News Coverage of Buddhist Monks in The Nation Online Newspaper: A Syntactic Analysis. It delves into syntactic sentence types, phrase-level structures in subjects and predicates, and word-level structures within the same context. The qualitative study utilizes data collected from The Nation's website over a decade (2013–2023) to analyze sentence types categorized into simple, compound, complex, and compound-complex, revealing a preference for simple (51.79%) and complex sentences (39.29%), indicating a journalistic inclination towards clarity and in-depth information presentation. Phrase-level analysis highlights a dominance of Noun Phrases (43.26%) and Prepositional Phrases (32.01%), underscoring their role in establishing relational context and subjects in sentences. Word-level analysis uncovers a significant use of determiners (44.07%) and nouns (38.45%), emphasizing their essential roles in conveying meaning and specifying noun phrases. The study's findings suggest a strategic balance in The Nation's syntactic choices, reflecting a nuanced approach to ensuring readability while providing comprehensive information, aligning with broader trends in digital journalism. This research contributes to understanding the syntactic strategies in journalism, particularly in the context of online media and its representation of religious figures.

Intelligent information extraction from government on-site inspection reports of construction projects: A graph-based text mining approach

Government inspection reports detail unsafe acts and conditions that arise on construction sites, especially front-line managers’ non-compliance practices, which are hardly identified during self-inspections. Such information serves as a valuable learning source for better construction management. However, non-compliance issue records in inspection reports are typically stored in unstructured text formats, making data analysis challenging. In response, an intelligent text mining framework integrating graph analysis and visualization is presented. The proposed framework comprises data collection and preprocessing and three levels of text analysis: word, sentence, and document. The main tasks of the word-level analysis include (1) extracting keywords using KeyBERT and (2) identifying non-compliance issue types based on community detection in a keyword co-occurrence graph. The sentence-level analysis is performed to automatically classify text data from inspection reports by determining the degree of similarity between texts and communities and assigning the most similar community to each text. The document-level analysis aims to identify the interrelations between various non-compliance issues through association rule mining and a community interaction network. The framework is validated by a total of 6,153 text data featuring non-compliance issues from 322 government on-site inspection reports in Shanghai, China. The results demonstrate that the critical word-level features of non-compliance issues can be accurately identified using KeyBert, which outperforms other state-of-the-art methods. Our approach can also automate the development of a data-driven taxonomy for non-compliance issues and the classification of the corresponding records, requiring less manual intervention than conventional text classification models.

A Comparison of Linguistic Aspects in 2015-2019 NAEA English Listening Texts Between School Levels and Between Language Forms

The current study aims to investigate English listening texts of the National Assessment of Educational Achievement (NAEA) at word, sentence, and text levels. It classifies NAEA listening texts into two school levels (middle school and high school) and two forms of spoken language (dialogues and monologues). It is necessary to examine changes of linguistic aspects in the NAEA for both middle and high school students to figure out whether text difficulty systematically rises as school level increases. The corpus for this study includes 169 listening texts from five academic years from 2015 to 2019. In this period, tests were implemented as paper-based tests. Two research questions are about linguistic differences of NAEA listening texts between middle school and high school and between dialogues and monologues at word, sentence and text levels. Coh-Metrix analyzed descriptive measures related to words, lexical diversity, and word information for word-level analysis, descriptive measures related to sentences and syntactic complexity for sentence-level analysis, and readability and referential cohesion for text-level analysis. Results and educational implications could increase the quality of NAEA listening texts.

A Hybrid Method of Long Short-Term Memory and Auto-Encoder Architectures for Sarcasm Detection

Sarcasm detection is considered one of the most challenging tasks in sentiment analysis and opinion mining applications in the social media. Sarcasm identification is therefore essential for a good public opinion decision. There are some studies on sarcasm detection that apply standard word2vec model and have shown great performance with word-level analysis. However, once a sequence of terms is being tackled, the performance drops. This is because averaging the embedding of each term in a sentence to get the general embedding would discard the important embedding of some terms. LSTM showed significant improvement in terms of document embedding. However, within the classification LSTM requires adding additional information in order to precisely classify the document into sarcasm or not. This study aims to propose two technique based on LSTM and Auto-Encoder for improving the sarcasm detection. A benchmark dataset has been used in the experiments along with several pre-processing operations that have been applied. These include stop word removal, tokenization and special character removal with LSTM which can be represented by configuring the document embedding and using Auto-Encoder the classifier that was trained on the proposed LSTM. Results showed that the proposed LSTM with Auto-Encoder outperformed the baseline by achieving 84% of f-measure for the dataset. The main reason behind the superiority is that the proposed auto encoder is processing the document embedding as input and attempt to output the same embedding vector. This will enable the architecture to learn the interesting embedding that have significant impact on sarcasm polarity.

Towards Automatic Collaboration Analytics for Group Speech Data Using Learning Analytics.

Collaboration is an important 21st Century skill. Co-located (or face-to-face) collaboration (CC) analytics gained momentum with the advent of sensor technology. Most of these works have used the audio modality to detect the quality of CC. The CC quality can be detected from simple indicators of collaboration such as total speaking time or complex indicators like synchrony in the rise and fall of the average pitch. Most studies in the past focused on “how group members talk” (i.e., spectral, temporal features of audio like pitch) and not “what they talk”. The “what” of the conversations is more overt contrary to the “how” of the conversations. Very few studies studied “what” group members talk about, and these studies were lab based showing a representative overview of specific words as topic clusters instead of analysing the richness of the content of the conversations by understanding the linkage between these words. To overcome this, we made a starting step in this technical paper based on field trials to prototype a tool to move towards automatic collaboration analytics. We designed a technical setup to collect, process and visualize audio data automatically. The data collection took place while a board game was played among the university staff with pre-assigned roles to create awareness of the connection between learning analytics and learning design. We not only did a word-level analysis of the conversations, but also analysed the richness of these conversations by visualizing the strength of the linkage between these words and phrases interactively. In this visualization, we used a network graph to visualize turn taking exchange between different roles along with the word-level and phrase-level analysis. We also used centrality measures to understand the network graph further based on how much words have hold over the network of words and how influential are certain words. Finally, we found that this approach had certain limitations in terms of automation in speaker diarization (i.e., who spoke when) and text data pre-processing. Therefore, we concluded that even though the technical setup was partially automated, it is a way forward to understand the richness of the conversations between different roles and makes a significant step towards automatic collaboration analytics.

Workforce diversity and quality improvement policies in early childhood education in East Asia

ABSTRACT This paper explores the relationship between workforce competence, migration, and quality improvement in the Asia-Pacific through analysis of Early Childhood Education (ECE) policy texts from three East Asia Summit countries. Content (word level) analysis of salient policy texts was undertaken to address two questions: How are culture and diversity conceptualized in these texts? How are diverse educators positioned in these texts? We show that, despite differing historical trajectories of ECE, policy frameworks in all three countries fail to sufficiently acknowledge the potential contribution of migrant educators to quality improvement initiatives. We conclude by calling for greater attention to workforce mobility and diversity in ECE policies, and for further research into the untapped competence of migrant educators.

Analysing Twitter Data for Phishing Tweets Identification

Detecting threats like adult, violent, and phishing tweets on online social networks is a crucial issue in recent years. The aim of the work is to identify phishing content from the users' perspective in real-time tweets. To outline such content comprehensively, lexicon analysis with sentiments are encapsulated to investigate tweets that yield phishing dynamic keywords, while some features and parameters are altered to optimize the performance. To support the preliminary study, the approach is rigorously designed to assemble users' opinions on completely different classes of phishing content. Each direct and indirect opinions as well as recently projected opinions are listed to characterize all sorts of phishing content. The authors use word level analysis with sentiments to build keyword blacklist lexicons. High promising results and high level of accuracy and performance are obtained experimentally if compared with the alternative algorithms.

Utilization of text mining as a big data analysis tool for food science and nutrition.

Big data analysis has found applications in many industries due to its ability to turn huge amounts of data into insights for informed business and operational decisions. Advanced data mining techniques have been applied in many sectors of supply chains in the food industry. However, the previous work has mainly focused on the analysis of instrument-generated data such as those from hyperspectral imaging, spectroscopy, and biometric receptors. The importance of digital text data in the food and nutrition has only recently gained attention due to advancements in big data analytics. The purpose of this review is to provide an overview of the data sources, computational methods, and applications of text data in the food industry. Text mining techniques such as word-level analysis (e.g., frequency analysis), word association analysis (e.g., network analysis), and advanced techniques (e.g., text classification, text clustering, topic modeling, information retrieval, and sentiment analysis) will be discussed. Applications of text data analysis will be illustrated with respect to food safety and food fraud surveillance, dietary pattern characterization, consumer-opinion mining, new-product development, food knowledge discovery, food supply-chain management, and online food services. The goal is to provide insights for intelligent decision-making to improve food production, food safety, and human nutrition.

A Sentence Repetition Task for Early Language Assessment in Spanish.

Sentence repetition tasks have been widely used in the last years as a diagnostic tool in developmental language disorders. However in Spanish there are few (if any) of these instruments, especially for younger children. In this context, we develop a new Sentence Repetition Task for assessing language (morphosyntactic) abilities of very young Spanish children. A list of 33 sentences of different length and complexity was created and included in the task. A total of 130 typical developing children from 2 to 4 years of age were engaged in a play situation and asked to repeat the sentences. Children's answers were scored for accuracy at sentence and word level and error analysis at the word level was undertaken. Besides a subsample of 92 children completed a non-word repetition task. First results show its adequacy to children from 2 to 4 years of age, its capacity to discriminate between different developmental levels, and its concurrent validity with the nonword repetition task.

Medical Corpora Comparison Using Topic Modeling

Abstract Free-form texts from electronic medical records are often used to build predictive models for medical and healthcare processes. In different medical centers, treatment of patients and other healthcare processes can occur in different ways according to the hospital’s internal protocols, which affects the structure of electronic medical records and the style of free-form texts. The paper aim is to compare two medical corpora in content to understand whether trained models of the first corpora apply to the second corpora. The approach contains topic modeling, topic segmentation, topic cross-segmentation and specific metric Topic Segmentation Collation to compare cross-segmentation results. Also, the results of the word-level analysis for both corpora are provided. We conclude each of the corpora needs different word-level processing and has a specific set of descriptions, which limits the use of predictive models for some diseases.

Understanding the language of entrepreneurship

Purpose Entrepreneurship is an activity with far-reaching economic and cultural implications. Research seeking to understand the cognition and behavior of entrepreneurs is devoting increasing attention to how entrepreneurs construct and utilize discourse. However, word-level analysis of the specific language used by entrepreneurs has not received significant attention. The purpose of this paper is to identify the words that comprise entrepreneurship discourse and describe how word-usage has changed over time. Design/methodology/approach To examine the language of entrepreneurship, the authors use modified MapReduce algorithms in conjunction with text mining techniques to construct a longitudinal corpus of approximately 2.5m words. The authors identify the most frequently used words in the entrepreneurship lexicon and then use content analysis to chart the evolution of word-use. Findings The findings reveal that entrepreneurs’ lexicon is complex and fluid. The most commonly used words suggest new trends in entrepreneurship. Originality/value The findings and methodological procedures contribute to research on entrepreneurs and the entrepreneurial process and, specifically, to work on entrepreneurial discourse, language-use and new venture communication. The findings also have implications for entrepreneurs and policymakers.

Word-level sentiment analysis with reinforcement learning

Reinforcement learning (RL) imitates how human perceive the word and acquire knowledge. We manage to propose a RL method to realize text sentiment analysis. We describe a new framework named Word-level Sentiment LSTM (WS-LSTM), which means we use such framework to get sentiment tendency for each word in a sentence. We suppose Positive, Neural and Negative as actions and establish three different LSTM tunnels for each action. When we choose an action for a word, relative LSTM tunnel will be chosen to handle the input. After traversing a whole sentence, we get both word-level sentiment sequence and a sentence-level representation. Such representation is used in classification and we success in getting a sentence-level sentiment analysis. Results show that our method can get sentiment for each word in a specific task. As to the word-level analysis, results from several datasets show that our method plays an acceptable job.

Tutorial: Assessment and Analysis of Polysyllables in Young Children.

Polysyllables, words of 3 or more syllables, represent almost 30% of words used in American English. The purpose of this tutorial is to support speech-language pathologists' (SLPs') assessment and analysis of polysyllables, extending the focus of published assessment tools that focus on sampling and analyzing children's segmental accuracy and/or the presence of phonological error patterns. This tutorial will guide SLPs through a review of 53 research papers that have explored the use of polysyllables in assessment, including the sampling and analysis procedures used in different research studies. The tutorial will also introduce two new tools to analyze and interpret polysyllable speech samples: the Word-Level Analysis of Polysyllables (WAP; Masso, 2016b) and the Framework of Polysyllable Maturity (Framework; Masso, 2016a). Connected speech and single-word sampling tasks were used across the 53 studies to elicit polysyllables, and a number of analysis methods were reported, including measures of segmental accuracy and measures of structural and suprasegmental accuracy. The WAP and the Framework extend SLPs' depth of analysis of polysyllables. SLPs need a range of clinical tools to support the assessment and analysis of polysyllables. A case study comparing different speech analysis methods demonstrates the clinical value in utilizing the WAP and the Framework to interpret children's polysyllable productions in addition to traditional methods of speech sampling and analysis.

Affective reasoning for big social data analysis

This special section focuses on the introduction, presentation, and discussion of novel techniques that further develop and apply affective reasoning tools and techniques for big social data analysis. A key motivation for this special section, in particular, is to explore the adoption of novel affective reasoning frameworks and cognitive learning systems to go beyond a mere word-level analysis of natural language text and provide novel concept-level tools and techniques that allow a more efficient passage from (unstructured) natural language to (structured) machine-processable affective data, in potentially any domain. The selected papers aim to address the wide spectrum of issues related to affective computing research and, hence, better grasp the current limitations and opportunities related to this fast-evolving branch of artificial intelligence. Out of the 29 submissions received, 5 were accepted to appear in the special section. One of the accepted papers underwent 3 rounds of revisions, the rest were revised twice. The papers appearing in this issue are briefly summarized.

A study on real-time low-quality content detection on Twitter from the users' perspective.

Detection techniques of malicious content such as spam and phishing on Online Social Networks (OSN) are common with little attention paid to other types of low-quality content which actually impacts users’ content browsing experience most. The aim of our work is to detect low-quality content from the users’ perspective in real time. To define low-quality content comprehensibly, Expectation Maximization (EM) algorithm is first used to coarsely classify low-quality tweets into four categories. Based on this preliminary study, a survey is carefully designed to gather users’ opinions on different categories of low-quality content. Both direct and indirect features including newly proposed features are identified to characterize all types of low-quality content. We then further combine word level analysis with the identified features and build a keyword blacklist dictionary to improve the detection performance. We manually label an extensive Twitter dataset of 100,000 tweets and perform low-quality content detection in real time based on the characterized significant features and word level analysis. The results of our research show that our method has a high accuracy of 0.9711 and a good F1 of 0.8379 based on a random forest classifier with real time performance in the detection of low-quality content in tweets. Our work therefore achieves a positive impact in improving user experience in browsing social media content.

Longitudinal changes in polysyllable maturity of preschool children with phonologically-based speech sound disorders

ABSTRACTChildren’s polysyllables were investigated for changes in (1) consonant and vowel accuracy, (2) error frequency and (3) polysyllable maturity over time. Participants were 80 children (4;0-5;4) with phonologically-based speech sound disorders who participated in the Sound Start Study and completed the Polysyllable Preschool Test (Baker, 2013) three times. Polysyllable errors were categorised using the Word-level Analysis of Polysyllables (WAP, Masso, 2016a) and the Framework of Polysyllable Maturity (Framework, Masso, 2016b), which represents five maturity levels (Levels A-E). Participants demonstrated increased polysyllable accuracy over time as measured by consonant and vowel accuracy, and error frequency. Children in Level A, the lowest level of maturity, had frequent deletion errors, alterations of phonotactics and alterations of timing. Participants in Level B were 8.62 times more likely to improve than children in Level A at Time 1. Children who present with frequent deletion errors may be less likely to improve their polysyllable accuracy.

Polysyllable productions in preschool children with speech sound disorders: Error categories and the Framework of Polysyllable Maturity

Purpose: Children with speech sound disorders (SSD) find polysyllables difficult; however, routine sampling and measurement of speech accuracy are insufficient to describe polysyllable accuracy and maturity. This study had two aims: (1) compare two speech production tasks and (2) describe polysyllable errors within the Framework of Polysyllable Maturity.Method: Ninety-three preschool children with SSD from the Sound Start Study (4;0–5;5 years) completed the Polysyllable Preschool Test (POP) and the Diagnostic Evaluation of Articulation and Phonology (DEAP-Phonology).Result: Vowel accuracy was significantly different between the POP and the DEAP-Phonology. Polysyllables were analysed using the seven Word-level Analysis of Polysyllables (WAP) error categories: (1) substitution of consonants or vowels (97.8% of children demonstrated common use), (2) deletion of syllables, consonants or vowels (65.6%), (3) distortion of consonants or vowels (0.0%), (4) addition of consonants or vowels (0.0%), (5) alteration of phonotactics (77.4%), (6) alteration of timing (63.4%) and (7) assimilation or alteration of sequence (0.0%). The Framework of Polysyllable Maturity described five levels of maturity based on children’s errors.Conclusions: Polysyllable productions of preschool children with SSD can be analysed and categorised using the WAP and interpreted using the Framework of Polysyllable Maturity.

Speech Sound Development of Toddlers in Spontaneous Speech: Segmental Level and Whole Word Level Analysis

말소리 장애 유아의 평가와 중재를 위해서는 일반 유아의 말소리 발달 양상에 대한 정보가 필요하다. 본 연구는 일반 유아의 말소리 발달 양상을 분절음 단위와 단어 단위 지표를 사용하여 종단적으로 살펴보는 것을 목적으로 하였다. 2세 유아 3명, 3세 유아 2명을 대상으로 1년 6개월의 기간 동안 평균 6개월(5-8개월) 간격으로 총 4회, 자발화 맥락에서 분절음 단위 지표로 음소목록 유형 수, 총 자음 산출 수, 자음정확도(PCC)를 평가하였고, 단어 단위 지표로 단어단위복잡률(PMLU), 단어단위근접률(PWP), 단어단위정확률(PWC)로 평가하였다. 종단적으로 확인한 유아들의 말소리 발달에서 몇 가지 특성을 확인할 수 있었다. 3세 후반에서 4세 초반에는 지표의 증가율이 둔화되고 일부 지표는 천장효과가 나타나 이 시기에 유아들의 말소리 발달이 완성기에 들어서는 것으로 추정할 수 있었다. 그러나 일부 유아는 3세 전반에 이미 그러한 양상을 보여주어 개인차를 보여 주었다. 평가 지표에 따른 차이도 나타났는데, 분절음 단위의 PCC는 3세 후반에서 평균 95%(90.3∼97.1%)에 도달한 반면, 단어 단위의 PMLU는 5세까지 지속적으로 변화하였다. 또한 PCC가 집단 평균을 따라가는 것으로 나타난 일부 유아는 PMLU에서 집단 평균에 도달하지 못하였다. 이는 취학 전 유아의 말소리 발달 평가에서 분절음과 단어 단위 분석결과를 종합하여 사용할 필요성이 있음을 시사한다.This study aims to examine developmental characteristics of speech sounds in toddlers through segmental and whole word evaluation. Three toddlers of 2 years old and two toddlers of 3 years old participated in the study for the duration of one and a half years. Evaluations of spontaneous speech were performed four times with 6 month interval. Three indexes for segment level such as Number of Phonetic Inventory, Total Number of Consonants, and Percentage of Correct Consonant(PCC) and 3 indexes for whole word level such as Phonological Mean Length of Utterance(PMLU), Proportion of Whole-Word Proximity(PWP), and Proportion of Whole-word Correctness(PWC) were used for evaluation. The longitudinal study revealed several characteristics of speech sound development of participants. Around the time when toddlers reach the age of 4, each index started to increase at a slower pace with some demonstrating ceiling effect. These results suggest that the phonological development of toddlers entered stable stage at this said period. That said. however, there were individual differences for each child. There were differences regarding indexes as well. On average, PCC reached 95% at the later period of the age 3, whereas PMLU continued to change until the age of 5. Some toddlers whose PCC was in the normal developmental range showed PMLU lower than the group average. These results suggest that a comprehensive use of both segmental and whole word level analyses may be necessary in evaluating the phonological development of toddlers.

Sentiment Data Flow Analysis by Means of Dynamic Linguistic Patterns

Abstract-Emulating the human brain is one of the core challenges of computational intelligence, which entails many key problems of artificial intelligence, including understanding human language, reasoning, and emotions. In this work, computational intelligence techniques are combined with common-sense computing and linguistics to analyze sentiment data flows, i.e., to automatically decode how humans express emotions and opinions via natural language. The increasing availability of social data is extremely beneficial for tasks such as branding, product positioning, corporate reputation management, and social media marketing. The elicitation of useful information from this huge amount of unstructured data, however, remains an open challenge. Although such data are easily accessible to humans, they are not suitable for automatic processing: machines are still unable to effectively and dynamically interpret the meaning associated with natural language text in very large, heterogeneous, noisy, and ambiguous environments such as the Web. We present a novel methodology that goes beyond mere word-level analysis of text and enables a more efficient transformation of unstructured social data into structured information, readily interpretable by machines. In particular, we describe a novel paradigm for real-time concept-level sentiment analysis that blends computational intelligence, linguistics, and common-sense computing in order to improve the accuracy of computationally expensive tasks such as polarity detection from big social data. The main novelty of the paper consists in an algorithm that assigns contextual polarity to concepts in text and flows this polarity through the dependency arcs in order to assign a final polarity label to each sentence. Analyzing how sentiment flows from concept to concept through dependency relations allows for a better understanding of the contextual role of each concept in text, to achieve a dynamic polarity inference that outperforms state-of-the-art statistical methods in terms of both accuracy and training time.

Sentic patterns: Dependency-based rules for concept-level sentiment analysis

The Web is evolving through an era where the opinions of users are getting increasingly important and valuable. The distillation of knowledge from the huge amount of unstructured information on the Web can be a key factor for tasks such as social media marketing, branding, product positioning, and corporate reputation management. These online social data, however, remain hardly accessible to computers, as they are specifically meant for human consumption. The automatic analysis of online opinions involves a deep understanding of natural language text by machines, from which we are still very far. To this end, concept-level sentiment analysis aims to go beyond a mere word-level analysis of text and provide novel approaches to opinion mining and sentiment analysis that enable a more efficient passage from (unstructured) textual information to (structured) machine-processable data. A recent knowledge-based technology in this context is sentic computing, which relies on the ensemble application of common-sense computing and the psychology of emotions to infer the conceptual and affective information associated with natural language. Sentic computing, however, is limited by the richness of the knowledge base and by the fact that the bag-of-concepts model, despite more sophisticated than bag-of-words, misses out important discourse structure information that is key for properly detecting the polarity conveyed by natural language opinions. In this work, we introduce a novel paradigm to concept-level sentiment analysis that merges linguistics, common-sense computing, and machine learning for improving the accuracy of tasks such as polarity detection. By allowing sentiments to flow from concept to concept based on the dependency relation of the input sentence, in particular, we achieve a better understanding of the contextual role of each concept within the sentence and, hence, obtain a polarity detection engine that outperforms state-of-the-art statistical methods.