Word Labels Research Articles

Zero-Shot Relation Classification Through Inference on Category Attributes.

The goal of relationship classification (RC) is to predict the semantic relationship between two entities in a given sentence. With the advent of deep learning and pretrained language models, RC research has progressed by leaps and bounds. However, the current studies are focused mainly on predicting semantic relationships from a predefined set. How to recognize unseen relationships remains a challenge, which is also known as the zero-shot RC (ZSRC) task. Some ZSRC-related methods directly map relationship categories to numerical indices, constraining the model's ability to autonomously infer and understand these relationships, while others rely heavily on manual definitions. To address these issues and inspired by the way of reasoning in which humans perform RC tasks, we propose a new framework to handle the ZSRC task through inference on category attributes (ICAs). The main idea of ICA is to detect the semantic relationship between promises, which are RC sentences, and hypotheses, which are relational sentences of entities created by templates. Specifically, instead of manual design, we introduce two hypothesis templates derived from the label words (LWs) and descriptions (LDs) associated with each relationship. These templates are used to automatically convert the RC data into the textual entailment (TE) format. Furthermore, they are fine-tuned with a pretrained TE model, facilitating the acquisition of relational knowledge and enabling the generalization of semantic reasoning rules learned from seen classes to unseen classes. Moreover, to implement multirelationship semantic inference for all unseen classes, we propose an entailment difference mechanism to enhance the reasoning capability of the model. Besides the current ZSRC test setting, we also examine our method in an even more challenging setting to deal with data scarcity in real-world applications. The outstanding performance of ICA on the FewRel and Wiki-ZSL datasets demonstrates its effectiveness in the ZSRC task.

Dual drivers of bilingual semantic accent: Semantic relations and input limitations

Research question: Second language (L2) learners may have a “semantic accent,” using L2 words differently from a native speaker. The current studies investigate how learning conditions and the need to use different semantic dimensions may constrain the acquisition of L2 meanings. Approach: In Experiment 1, we asked whether first language (L1) Mandarin speakers would successfully learn English carry and hold with intensive input in a laboratory training procedure. In Experiment 2, we asked whether the difficulties observed were due to differing semantic dimensions per se or to the need to move from a larger number of fine-grained native distinctions to a smaller number of broader English meanings. Data and analysis: Word definitions and label choice after training were scored for consistency with native performance. Conclusions: Experiment 1 supported the idea that intensive input helps, but Experiment 2 suggested that any kind of semantic difference from L1 is challenging even with intensive input. Originality: The current study clarifies the contributions of learning conditions and L1-L2 semantic differences in semantic accent. Implications: Reducing semantic accent may require instruction in specific L1-L2 differences and cultivation of metacognitive sensitivity to the potential for differences.

Improving consumer understanding of pesticide toxicity labels: experimental evidence

Consumers often inadvertently misperceive the health hazards associated with over-the-counter pesticides under the current textual labeling policy, potentially leading to improper use. We conducted an incentivized framed field experiment with eye tracking to evaluate the effectiveness of the current pesticide labels that convey risk using signal words (Caution, Warning, Danger) compared to two visually focused label alternatives: traffic light colors and skull intensity symbols. A total of 166 participants were randomly assigned to one of three label formats and asked to rank toxicity levels and make purchasing decisions within multiple price lists. Results show that signal words fail to adequately communicate toxicity levels. Specifically, participants’ correct assessment of toxicity level dramatically improves from 54% under the existing signal word label to 95% under the traffic light and 83% under the skull intensity symbol labels. We also find that participants are more likely to choose the less toxic alternatives under the new labels, suggesting the current labeling system may affect choice and have unintended adverse effects on human health.

The Relationship Between Subjective Memory Experience and Objective Memory Performance Remains Stable Across the Lifespan

The way humans remember events changes across the lifespan. Older adults often rate the vividness of their memories as being greater or equal to younger adults, despite poorer performance on episodic memory tasks. This study explored how the content (place, person and object) and specificity (conceptual gist versus perceptual detail) of event memories relate to the subjective experience of memory vividness and memory confidence, and how this relationship is affected by healthy ageing. 100 healthy older adults and 100 young adults were tested online, using an adapted version of a paradigm developed by Cooper and Ritchey (2022). At encoding, participants generated a distinctive story to associate together (1) a theme word, and images of (2) a famous person, (3) a place, and (4) an object, to create unique events. At test, participants identified the event components using word labels (indexing conceptual gist), and the studied images (indexing perceptual details). Replicating Cooper and Ritchey (2022), we found that young adults’ memory vividness ratings were related to their memory for the conceptual gist of the events, with no modulation by the type of the content recalled. Strikingly, older adults showed the same relationship between vividness measures and objective performance as the young adults. Contrary to some previous studies, we found that older adults obtained lower scores for gist-based memory, and their vividness ratings were correspondingly lower than the younger adults. Across both age groups, vividness and confidence ratings followed a similar pattern, showing a stronger relationship with conceptual gist. Our results suggest that throughout the lifespan, the amount of conceptual information retrieved about an event relates to the ability to reexperience it vividly, and to have confidence in one’s memory.

Semantic context-dependent neural representations of odors in the human piriform cortex revealed by 7T MRI.

Olfactory perception depends not only on olfactory inputs but also on semantic context. Although multi-voxel activity patterns of the piriform cortex, a part of the primary olfactory cortex, have been shown to represent odor perception, it remains unclear whether semantic contexts modulate odor representation in this region. Here, we investigated whether multi-voxel activity patterns in the piriform cortex change when semantic context modulates odor perception and, if so, whether the modulated areas communicate with brain regions involved in semantic and memory processing beyond the piriform cortex. We also explored regional differences within the piriform cortex, which are influenced by olfactory input and semantic context. We used 2 × 2 combinations of word labels and odorants that were perceived as congruent and measured piriform activity with a 1-mm isotropic resolution using 7T MRI. We found that identical odorants labeled with different words were perceived differently. This labeling effect was observed in multi-voxel activity patterns in the piriform cortex, as the searchlight decoding analysis distinguished identical odors with different labels for half of the examined stimulus pairs. Significant functional connectivity was observed between parts of the piriform cortex that were modulated by labels and regions associated with semantic and memory processing. While the piriform multi-voxel patterns evoked by different olfactory inputs were also distinguishable, the decoding accuracy was significant for only one stimulus pair, preventing definitive conclusions regarding the locational differences between areas influenced by word labels and olfactory inputs. These results suggest that multi-voxel patterns of piriform activity can be modulated by semantic context, possibly due to communication between the piriform cortex and the semantic and memory regions.

DXVNet-ViT-Huge (JFT) Multimode Classification Network Based on Vision Transformer

Aiming at the problem that traditional CNN network is not good at extracting global features of images, Based on DXVNet network, Conditional Random Fields (CRF) component and pre-trained ViT-Huge (Vision Transformer) are adopted in this paper Transformer model expands and builds a brand new DXVNet-ViT-Huge (JFT) network. CRF component can help the network learn the constraint conditions of each word corresponding prediction label, improve the D-GRU method based word label prediction errors, and improve the accuracy of sequence annotation. The Transformer architecture of the ViT (Huge) model can extract the global feature information of the image, while CNN is better at extracting the local features of the image. Therefore, the ViT (Huge) Huge pre-training model and CNN pre-training model adopt the multi-modal feature fusion technology. Two complementary image feature information is fused by Bi-GRU to improve the performance of network classification. The experimental results show that the newly constructed Dxvnet-Vit-Huge (JFT) model achieves good performance, and the F1 values in the two real public data sets are 6.03% and 7.11% higher than the original DXVNet model, respectively.

Dzongkha Next Words Prediction Using Bidirectional LSTM

Dzongkha Development Commission of Bhutan (DDC) is trying to computerize Dzongkha. However, the computerization of Dzongkha poses numerous challenges. Currently, the support for Dzongkha in modern technology is limited to printing, typing, and storage. Typewriting a single Dzongkha word requires several keypresses. As a result, typing Dzongkha is tedious. In this paper, the Dzongkha word label prediction was studied. The purpose of the study was to further reduce keystrokes and make Dzongkha typing much faster. The dataset encompasses different genres curated by DDC. The dataset consisted of 10000 sentences and 4820 unique words. Next, 52150 sequences were generated using N-gram methods followed by vectorizing text using embedding techniques. Different RNN-based models were evaluated for the next Dzongkha words prediction. Two Bi-LSTM layers with 512 hidden layer neurons gave the best accuracy of 73.89% with a loss of 1.0722.

Linguistic, historical, and religious connotations of the term "Hebrew" in Jewish, Christian, and Islamic sources

There is no doubt that the Hebrew language, which is an important part of the Semitic languages to which our Arabic language belongs, has become one of the important languages in the Arab region, and it is time for Arabs to learn it and understand what is going on around them and from within them in terms of plans that reflect on the future of the region as a whole. Therefore, it was our duty as specialists in the study of Semitic languages and the Hebrew language in particular to write research on the origins and divisions of this language. From this point of view, this research was prepared to complement the great scientific research and studies in the field of Semitic studies, especially the Hebrew side of it, the research has been addressed to several topics, the most important of which was the definition of the term "Hebrew" in Arabic and Hebrew, and knowing the origin of this term with mentioning the most important opinions that touched on the hadith about its origin and interpretation, and also the origin of the word labels (Jews - Israelites - Hebrews), and then it was proven whether the words "Hebrew" and "Arabic" are one word? Or is there a difference between them, hence the historical and religious connotations of the term "Hebrew" are mentioned in Jewish sources, and then in Jewish sources and finally in Islamic sources, where these connotations are the main axes around which this research revolves. In the end, there was the conclusion in which we mentioned the most important results we reached through the preparation of this research.

Adaptive Local Context and Syntactic Feature Modeling for Aspect-Based Sentiment Analysis

Aspect-based sentiment analysis is a fine-grained sentiment analysis task that consists of two types of subtasks: aspect term extraction and aspect sentiment classification. In the aspect term extraction task, current methods suffer from the lack of fine-grained information in aspect term extraction and difficulty in identifying aspect term boundaries. In the aspect sentiment classification task, the current aspect sentiment classifier cannot adapt itself to the text and determine the local context. To address these two challenges, this work proposes an adaptive semantic relative distance approach based on dependent syntactic analysis, which uses adaptive semantic relative distance to determine the appropriate local context for each text and increase the accuracy of sentiment analysis. Meanwhile, the study also predicts the current word labels by combining local information features extracted by local convolutional neural networks and global information features to precisely locate the word labels. In two subtasks, our proposed model improves accuracy and F1 scores on the SemEval-2014 Task 4 Restaurant and Laptop datasets compared to the state-to-the-art approaches, especially in the aspect sentiment classification subtask.

Automated Voice-to-Image Generation Using Generative Adversarial Networks in Machine Learning

Creating visuals from words may appear to be a complex process, but it is achievable with today’s technological advancements in Information Systems. Naturally, all the human-centric actions and assumptions may lead to visualization using Artificial Intelligence. In today’s Information Systems technological world, any item or a thing can be best described in pictorial form as a human person. Our paper aims to focus on providing machines with this intelligence. To complete this challenge, we used Natural Language Processing with Deep Learning. Our primary focus is on Generative Adversarial Networks. GANs will generate data based on word labels that are provided. NLP is also important since it helps to translate the provided speech into embedding vectors that the model can use. Our study is on the CUB dataset, which comprises bird photos. In today’s world, there are text-to-image generating models accessible. The authors investigated all of them, extending text-to-image generation to voice-to-image generation.

Speech and social cues combine at discourse boundaries to promote word learning

Social cues, such as eye gaze and referential gestures, have been shown to promote word learning in children, usually by exposing children to a label-object association with vs. without a social cue. However, social cues are not always so cleanly linked to referents in the world. Instead, label-object associations occur imperfectly across time in a discourse. We conducted an experiment to examine how labels and eye gaze interactively support 4- and 5-year-olds’ word learning during exposures to a series of brief discourse topics. We found that children learned novel words better when social cues and word labels jointly marked the beginning or end of a discourse topic, but not when they were misaligned. The findings indicate that young learners can use social cues to inform their understanding of word meanings across successive utterances, not just within single labeling events. We discuss how social cues may support children’s ability to find where discourse topics begin and end within their ongoing speech input.

POS Tagger Improvisation with the Addition of Foreign Word Labels on Telkom University News

News is a medium of daily information usually obtained by the public. The news consists of a lot of information in it and is composed of sentence structures. Each language is unique with its own sentence structure, like Indonesian and other foreign languages. But nowadays, many media mix Indonesian with foreign languages, making the sentence structure different from Bahasa Indonesia. To classify these words, Part Of Speech Tagging needed to determine the class of words composed of sentences by learning from the Corpus of each language. With the new sentence structure, POS Tagger requires a larger Corpus to learn. The language structure can determine the results of tagging from the POS Tagger. If there are words that are not in the Corpus, it can reduce the accuracy of the POS Tagger. We conducted to enhance the research results by adding data with a different sentence structure from the Indonesian Language Corpus using sentences from online media. Added about 242 sentences with 7,043 tokens on Corpus focused on Foreign Word tags, which total 3819 tags. After doing some testing and scenarios, the results of the accuracy of POS Tagger show an accuracy of 94.7% using the Hidden Markov Model method with the F1-Score tag FW 78%.

Research on Regional Energy Visualization Based on Knowledge Graph

The equipment information in the regional energy system is difficult to retrieve and the amount of data is too large to realize intelligent application. In order to solve the problem, this paper proposes a regional energy visualization research method based on knowledge graph. Firstly, the collected regional energy information is classified. Secondly, the framework is defined for the unstructured data, which is difficult to be recognized by computer, and the hidden Markov model is used for word segmentation and labeling. Finally, the regional energy equipment information data are processed to construct the knowledge graph, the regional energy map structure information is represented by the ternary group graph structure, and the regional energy retrieval system is designed. The system can efficiently process and retrieve a large amount of natural language information in regional energy and realize the intelligence of regional energy equipment. It provides a strong guarantee for the efficient and stable operation of regional energy-related equipment and has a wide range of practical application and research value.

Darwin's tales-A content analysis of how evolution is presented in children's books.

In science, certain theories led to a paradigm shift in human being’s approach to explain nature, such as the theory of relativity, the quantum theory, and the theory of evolution. The latter explains the emergence of biodiversity on Earth and all living beings’ relatedness, including humans. Accordingly, evolutionary theory is a central part of scientific literacy. However, scholars have demonstrated that misconceptions emerging in childhood hinder learners from grasping evolutionary processes. Implementing evolution in early science education could enhance scientific ideas as a basis for subsequent learning at school. Currently, children’s literature that deals with evolution is increasing and may enable more children to encounter evolutionary theory before entering school. This explorative study aimed to analyze how children’s books about evolution approach explaining this complex topic to young children in terms of covered contents, underlying concepts and use of language. We conducted (1) a text-based qualitative content analysis of 31 children’s books in the categories of organismal context, evolutionary principles, and misconceptions, and (2) a computer-supported content analysis of 33 word labels concerning (a) scientific terms and (b) verbs expressing evolutionary change. Although evolution is a universal concept, children’s books seem to promote specific contexts such as animal and human evolution. Even though the principle of selection requires an understanding of complex interactions between individuals and environmental factors, this principle was more frequent than the principles variation and inheritance. Phylogenetic history was covered more often than basic evolutionary processes, and evolutionary change was mainly mentioned at the species level over long periods. Besides, most books conveyed misconceptions such as transformationist, teleological or anthropomorphic reasoning. Consequently, books covering evolution may bias children’s first ideas concerning this topic or introduce unscientific ideas. Based on our results, we propose implications for early evolution educators and education researchers.

Long Song Type Classification based on Lyrics

Mongolian folk songs are inspired by Mongolian labor songs and are classified into long and short songs. Mongolian long songs have ancient origins, are rich in legends, and are a great source of folklore. So it was inscribed by UNESCO in 2008. Mongolian written literature is formed under the direct influence of oral literature. Mongolian long song has 3 classes: ayzam, suman, and besreg by their lyrics and structure. In ayzam long song, the world perfectly embodies the philosophical nature of world phenomena and the nature of human life. Suman long song has a wide range of topics such as the common way of life, respect for ancestors, respect for fathers, respect for mountains and water, livestock and animal husbandry, as well as the history of Mongolia. Besreg long songs are dominated by commanded and trained characters. In this paper, we proposed a method to classify their 3 types of long songs using machine learning, based on their lyrics structures without semantic information. We collected lyrics of over 80 long songs and extracted 11 features from every single song. The features are the name of a song, number of the verse, number of lines, number of words, general value, double value, elapsed time of verse, elapsed time of 5 words, and the longest elapsed time of 1 word, full text, and type label. In experimental results, our proposed features show on average 78% recognition rates in function type machine learning methods, to classify the ayzam, suman, and besreg classes.

English Lexical Analysis System of Machine Translation Based on Simple Recurrent Neural Network.

With the increasing communication between countries, machine translation has become an important way of communication between ethnic groups in different language systems. In order to further improve the coverage mechanism and vocabulary translation quality of the machine translation model of neural network, this paper will study the machine translation English lexical analysis system based on simple recurrent neural network. In order to improve the effect of word alignment, marking and attention mechanism are introduced into the English lexical analysis model. The experimental results show that the tagging of named entity words in the system can reduce the problem of unregistered words. Combined with the attention mechanism, it can significantly improve the effect of word alignment and label recall. It not only improves the controllability of translation but also has a positive impact on the quality of translation and its application effect in specific scenes. The evaluation of translation quality indicators shows that the system can effectively improve the accuracy and quality of Chinese-English translation.

INTELLECTUAL PROPERTY RIGHT - TRADEMARK: GENERAL REVIEW

An Act to amend and consolidate the law relating to trademarks, to provide for registration and better protection of trademarks for goods and services and for the prevention of the use of fraudulent marks. A trademark (popularly known as brand name) in layman's language is a visual symbol which may be a word signature, name, device, label, numerals or combination of colours used by one undertaking on goods or services or other articles of commerce to distinguish it from other similar goods or services originating from a different undertaking.

Word recognition and automated epenthesis removal for Indonesian sign system sentence gestures

This research <span>focuses on building a system to translate continuous Indonesian sign system (SIBI) gestures into text. In a continuous gesture, a signer will add an epenthesis (transitional) gesture, which is hand movement with no meaning but needed to connect the hand movement of one word with the next word in a continuous gesture. Reducing the number of irrelevant inputs to the model through automated epenthesis removal can improve the system's ability to recognize the words in continuous gestures. We implemented threshold conditional random fields (TCRF) to identify epenthesis gestures. The dataset consists of 2,255 videos representing 28 common sentences in SIBI. The translation system consists of MobileNetV2 as a feature extraction technique, removing epenthesis gestures found by the TCRF, and a long short-term memory (LSTM) for the classifier. With the MobileNetV2-TCRF-bidirectional LSTM model, the best word error rate (WER) and sentence accuracy (SAcc) were 33.4% and 16.2%, respectively. Intermediate-stage processing steps consisting of sandwiched majority voting of the TCRF and the removal of word labels whose number of frames is less than two frames, along with LSTM output grouping, were able to reduce WER from 33.4% to 3.4% and increase SAcc from 16.2% to 80.2%.</span>

A novel textual data augmentation method for identifying comparative text from user-generated content

• A framework is proposed for comparative text identification from user-generated content. • A syntax-aware and sentiment polarity-aware text augmentation method is proposed. • A series of experiments verify the feasibility and effectiveness of the framework. Mining user-generated content on e-commerce platforms and social media is timely and more objective compared with other information access channels for gaining competitive intelligence. Identifying comparative text from large volumes of non-comparative text is an important but challenging task. On one hand, existing methods are time-consuming and not generalizable across different domains. On the other hand, the datasets for the task generally suffer from the severe imbalance issue. To address abovementioned problems, we propose a framework adopting advanced deep learning methods to automatically learn features and a novel textual data augmentation method named TA3S to deal with the data imbalance issue. Specifically, the TA3S method simultaneously considers the syntactic structure and semantic information of comparative text samples. Moreover, in order to support the successful implementation of TA3S, we develop a novel method based on word embedding and label propagation algorithm to distinguish between synonymous and antonymous substitute words. The experiments on two real-world datasets demonstrate the feasibility and effectiveness of our framework, and present that our framework outperforms state-of-the-art methods in identifying comparative text from user-generated content.

Verbal descriptors of the frequency of side effects: implementation of EMA recommendations in patient information leaflets in Poland.

The European Medicines Agency (EMA) recommends a description of drug side effects based on the frequency format and the associated verbal description. Although the recommendations refer to English-speaking countries, in several non-English-speaking states, official authorities have accepted the proposed recommendations on how the patient information leaflets should be designed for descriptions of side effect frequencies. The aim of the study was to examine how manufacturers of authorized medicines in Poland implement the EMA recommendations regarding the verbal descriptors of the frequency of side effects. A qualitative study. As a first step, we identified which of the 150 pharmaceutical companies operating in Poland had the largest market share. Then, five manufacturers were selected at random from the list of the top 15 drug manufacturers in Poland by market share of the pharmaceutical sector. Lists of medicinal products authorized for marketing in Poland were downloaded from manufacturers' websites, and then five products from each manufacturer were selected based on random sampling. The study included only prescription medicines and excluded over-the-counter medicines and dietary supplements from the sample. Subsequently, for each of the 25 drugs, relevant patient information leaflets were obtained from the manufacturers' websites. We evaluated how information on the frequency of side effects was provided in each leaflet, including the use of EMA-recommended terms (verbal descriptors such as 'very common,' 'common,' 'uncommon,' 'rare' and 'very rare') and additional notes explaining their meanings. For all manufacturers, word labels of the frequencies of side effects selected for the study were the same, but the additional notes explaining their meanings were different. There were various explanations of how to understand verbal descriptors of the frequency of side effects not only across different manufacturers but also across different medicines from one manufacturer. There is no single standard in the Polish pharmaceutical industry for implementing the EMA recommendations into the written information about the frequency of side effects. The observed differences for an explanation of how to understand a given verbal term do not favor a uniform interpretation of the verbal frequency labels meaning by patients.