Semantic Similarity Research Articles

Extracting Sentence Embeddings from Pretrained Transformer Models

Pre-trained transformer models shine in many natural language processing tasks and therefore are expected to bear the representation of the input sentence or text meaning. These sentence-level embeddings are also important in retrieval-augmented generation. But do commonly used plain averaging or prompt templates sufficiently capture and represent the underlying meaning? After providing a comprehensive review of existing sentence embedding extraction and refinement methods, we thoroughly test different combinations and our original extensions of the most promising ones on pretrained models. Namely, given 110 M parameters, BERT’s hidden representations from multiple layers, and many tokens, we try diverse ways to extract optimal sentence embeddings. We test various token aggregation and representation post-processing techniques. We also test multiple ways of using a general Wikitext dataset to complement BERT’s sentence embeddings. All methods are tested on eight Semantic Textual Similarity (STS), six short text clustering, and twelve classification tasks. We also evaluate our representation-shaping techniques on other static models, including random token representations. Proposed representation extraction methods improve the performance on STS and clustering tasks for all models considered. Very high improvements for static token-based models, especially random embeddings for STS tasks, almost reach the performance of BERT-derived representations. Our work shows that the representation-shaping techniques significantly improve sentence embeddings extracted from BERT-based and simple baseline models.

Development of a Children’s Educational Dictionary for a Low-Resource Language Using AI Tools

Today, various interactive tools or partially available artificial intelligence applications are actively used in educational processes to solve multiple problems for resource-rich languages, such as English, Spanish, French, etc. Unfortunately, the situation is different and more complex for low-resource languages, like Kazakh, Uzbek, Mongolian, and others, due to the lack of qualitative and accessible resources, morphological complexity, and the semantics of agglutinative languages. This article presents research on early childhood learning resources for the low-resource Kazakh language. Generally, a dictionary for children differs from classical educational dictionaries. The difference between dictionaries for children and adults lies in their purpose and methods of presenting information. A themed dictionary will make learning and remembering new words easier for children because they will be presented in a specific context. This article discusses developing an approach to creating a thematic children’s dictionary of the low-resource Kazakh language using artificial intelligence. The proposed approach is based on several important stages: the initial formation of a list of English words with the use of ChatGPT; identification of their semantic weights; generation of phrases and sentences with the use of the list of semantically related words; translation of obtained phrases and sentences from English to Kazakh, dividing them into bigrams and trigrams; and processing with Kazakh language POS pattern tag templates to adapt them for children. When the dictionary was formed, the semantic proximity of words and phrases to the given theme and age restrictions for children were taken into account. The formed dictionary phrases were evaluated using the cosine similarity, Euclidean similarity, and Manhattan distance metrics. Moreover, the dictionary was extended with video and audio data by implementing models like DALL-E 3, Midjourney, and Stable Diffusion to illustrate the dictionary data and TTS (Text to Speech) technology for the Kazakh language for voice synthesis. The developed thematic dictionary approach was tested, and a SUS (System Usability Scale) assessment of the application was conducted. The experimental results demonstrate the proposed approach’s high efficiency and its potential for wide use in educational purposes.

Expert-Guided Large Language Models for Clinical Decision Support in Precision Oncology.

Rapidly expanding medical literature challenges oncologists seeking targeted cancer therapies. General-purpose large language models (LLMs) lack domain-specific knowledge, limiting their clinical utility. This study introduces the LLM system Medical Evidence Retrieval and Data Integration for Tailored Healthcare (MEREDITH), designed to support treatment recommendations in precision oncology. Built on Google's Gemini Pro LLM, MEREDITH uses retrieval-augmented generation and chain of thought. We evaluated MEREDITH on 10 publicly available fictional oncology cases with iterative feedback from a molecular tumor board (MTB) at a major German cancer center. Initially limited to PubMed-indexed literature (draft system), MEREDITH was enhanced to incorporate clinical studies on drug response within the specific tumor type, trial databases, drug approval status, and oncologic guidelines. The MTB provided a benchmark with manually curated treatment recommendations and assessed the clinical relevance of LLM-generated options (qualitative assessment). We measured semantic cosine similarity between LLM suggestions and clinician responses (quantitative assessment). MEREDITH identified a broader range of treatment options (median 4) compared with MTB experts (median 2). These options included therapies on the basis of preclinical data and combination treatments, expanding the treatment possibilities for consideration by the MTB. This broader approach was achieved by incorporating a curated medical data set that contextualized molecular targetability. Mirroring the approach MTB experts use to evaluate MTB cases improved the LLM's ability to generate relevant suggestions. This is supported by high concordance between LLM suggestions and expert recommendations (94.7% for the enhanced system) and a significant increase in semantic similarity from the draft to the enhanced system (from 0.71 to 0.76, P = .01). Expert feedback and domain-specific data augment LLM performance. Future research should investigate responsible LLM integration into real-world clinical workflows.

ANTONYMIC NESTS IN TOPONYMIC SYSTEM OF MODERN RUSSIAN

In the approach to solving the problem of antonymy, two aspects of its study have been outlined. Antonymy can only be considered within the framework of linguistic phenomena, and in this case, the origins of antonymy lie in the features of the functioning of linguistic units in the process of communication, that is, in speech activity. On the other hand, antonymy can be considered in the aspect of the logical-semantic correlation of concepts denoted by linguistic units. A special place in the Russian language is occupied by antonyms -words that are opposite in meaning. Antonymy reflects an essential aspect of systemic relations in Russian vocabulary. Modern linguistics considers synonymy and antonymy as extreme and borderline cases of content interchangeability and opposition of words. While synonymous relations are characterized by semantic similarity, antonymic relations are characterized by semantic difference. The existenceof antonyms in language lies in the nature of our perception of reality in all its contradictory complexity, in the unity and struggle of opposites. Therefore, contrasting words, as well as the concepts they denote, are not only opposites, but also closely related.

Towards Frugal Industrial AI: a framework for the development of scalable and robust machine learning models in the shop floor

AbstractArtificial intelligence (AI) among other digital technologies promise to deliver the next level of process efficiency of manufacturing systems. Although these solutions such as machine learning (ML) based condition monitoring and quality inspection are becoming popular, these work under very limited conditions. Solutions do not scale-up in the real environment, where there is a mix of manufacturing equipment, where the quality and quantity of data available changes from machine to machine, or where the process changes, changing the distribution of data (i.e. concept drift). This is particularly challenging in highly reconfigurable and flexible environments. Having to develop machine learning models from scratch every single time is not cost-effective, time-consuming, requires expert knowledge that is typically not available in the manufacturing environment as well as can be challenging when data is not available in high volumes. Model robustness, reusability, adaptability and life cycle management are the keys to scale-up this technology in the manufacturing industry. In this work, a conceptual framework to enable simple and robust ML model development for the shop floor is introduced. Referred here as Frugal Industrial AI, the approach takes advantage of existing models and their context to build more robust ones in a data-efficient manner. Using a semantic knowledge base of how to construct these models for different manufacturing applications and semi-automating the development or reuse of solutions through semantic similarity, it is demonstrated how models can be developed in a more streamlined way. In addition, it is demonstrated how capturing process context information is important for the effective reuse of existing models through continual learning. This is key to building more robust ML solutions that can deal with real changing manufacturing environments, avoiding retraining from scratch as well as enabling the non-expert to use AI effectively on the shop floor.

A Fuzzy Multigranularity Convolutional Neural Network With Double Attention Mechanisms for Measuring Semantic Textual Similarity

A Fuzzy Multigranularity Convolutional Neural Network With Double Attention Mechanisms for Measuring Semantic Textual Similarity

Enhancing Keyword Extraction from Academic Articles Using Highlights

ABSTRACTKeywords facilitate rapid comprehension of academic papers for scholars, enhancing research efficiency. As some papers lack author‐assigned keywords, automated keyword extraction becomes crucial. Addressing the limited utilization of external paper information beyond titles and abstracts in prior studies, this research proposed leveraging the highlights section, which summarizes contributions and novelties, for unsupervised keyword extraction. We demonstrated performance improvements across three unsupervised keyword extraction models by integrating highlight information into abstracts. Further analysis indicates we can filter out irrelevant abstract sentences based on their semantic similarity to the highlights. Overall, this study pioneers the exploration of utilizing paper highlights for keyword extraction, simultaneously boosting unsupervised keyword extraction performance and providing insights for other abstract‐based text mining applications.

A Rose by Another Name? Odor Misnaming is Associated with Linguistic Properties.

Naming common odors is a surprisingly difficult task: Odors are frequently misnamed. Little is known about the linguistic properties of odor misnamings. We test whether odor misnamings of old adults carry information about olfactory perception and its connection to lexical-semantic processing. We analyze the olfactory-semantic content of odor source naming failures in a large sample of older adults in Sweden (n = 2479; age 58-100 years). We investigate whether linguistic factors and semantic proximity to the target odor name predict how odors are misnamed, and how these factors relate to overall odor identification performance. We also explore the primary semantic dimensions along which misnamings are distributed. We find that odor misnamings consist of surprisingly many vague and unspecific terms, such as category names (e.g., fruit) or abstract or evaluative terms (e.g., sweet). Odor misnamings are often strongly associated with the correct name, capturing properties such as its category or other abstract features. People are also biased toward misnaming odors with high-frequency terms that are associated with olfaction or gustation. Linguistic properties of odor misnamings and their semantic proximity to the target odor name predict odor identification performance, suggesting that linguistic processing facilitates odor identification. Further, odor misnamings constitute an olfactory-semantic space that is similar to the olfactory vocabulary of English. This space is primarily differentiated along pleasantness, edibility, and concreteness dimensions. Odor naming failures thus contain plenty of information about semantic odor knowledge.

Multi-view graph contrastive learning for social recommendation.

With the rapid popularity of online social media, recommendation systems have increasingly harnessed social relations to enhance user-item interactions and mitigate the data sparsity issue. Beyond social connections, the semantic relatedness among items has emerged as a crucial factor in comprehending their inherent connections. In this work, we propose a novel Multi-view Contrastive learning framework for Social Recommendation, named MultiCSR. This framework adaptively incorporates user social networks and item knowledge graphs into modeling users preferences within recommendation systems. To facilitate the alignment of different views, we introduce a dedicated multi-view contrastive learning process that extracts rich information from each view and foster mutual enhancement. Extensive experiments conducted on three real-world datasets demonstrate the effectiveness of our framework over representative recommendation methods. Furthermore, ablation studies offer a deeper understanding of the mechanisms underlying our framework.

The Semantics of Happiness and Joy in the Italian Language

Abstract A number of contemporary researchers have proposed considering happiness and joy as distinct phenomena and theoretical constructs. However, there is as yet insufficient information regarding whether the theoretical and empirical arguments for distinguishing happiness from joy also apply to folk representations of these concepts. This work presents two studies in which Italian speakers performed a free-listing task asking them to list words semantically equivalent to the Italian words felicità (happiness) (Study 1) and gioia (joy) (Study 2). A comparison between the two studies revealed that the Italian concepts of happiness and joy had 49 lexical items in common, 32 of which related to emotions. Interestingly, 11 of the latter ranked in the top 10 cognitive salience indexes obtained from both Study 1 and Study 2. It is argued that the semantic similarities between the two words examined – felicità and gioia – should not overshadow the meaning nuances distinguishing them.

A Novel Optimization Scheme for Named Entity Recognition with Pre-trained Language Models

Named Entity Recognition (NER) is crucial for extracting structured information from text. While traditional methods rely on rules, Conditional Random Fields (CRFs), or deep learning, the advent of large-scale Pre-trained Language Models (PLMs) offers new possibilities. PLMs excel at contextual learning, potentially simplifying many natural language processing tasks. However, their application to NER remains underexplored. This paper investigates leveraging the GPT-3 PLM for NER without fine-tuning. We propose a novel scheme that utilizes carefully crafted templates and context examples selected based on semantic similarity. Our experimental results demonstrate the feasibility of this approach, suggesting a promising direction for harnessing PLMs in NER.

LongT5Rank: A Novel Integrated Hybrid Approach for Text Summarisation

Text summarisation reduces text length while retaining important information, helping individuals, especially students, in managing information overload during research or assignments. However, existing text summarisation methods often lose important details, generate irrelevant or redundant sentences, or produce incoherent summaries. This study introduces a hybrid approach, LongT5Rank (coined in this study), which combines TextRank, an extractive summarisation algorithm, with LongT5, an abstractive summarisation algorithm, to automate the summarisation process. TextRank utilizes GloVe, a pre-trained word embedding model, and PageRank, a graph-based ranking algorithm, to select representative sentences. LongT5, an encoder-decoder transformer model, condenses extracted sentences into a concise and coherent summary, handling input sequences up to 16,384 tokens, for long-range sequence-to-sequence tasks. The LongT5Rank approach has shown significant achievements, including a minimum 60% compression rate, a minimum 0.6 semantic textual similarity score, and an improved F-measure compared to employing TextRank alone. Furthermore, it received positive feedback from Human Level Performance (HLP), underlining the importance of evaluating results directly from human users. This emphasizes the belief that the performance of the proposed solution should be assessed natively by humans. By combining both extractive and abstractive methods, LongT5Rank excels in generating accurate and coherent summaries.

Phenotypic spectrum of dual diagnoses in developmental disorders

As more patients receive genome-wide sequencing, the number of individuals diagnosed with multiple monogenic conditions is increasing. We sought to investigate the relative phenotypic contribution of dual diagnoses using both manual curation and computational approaches. First, we computed 1,003,236 semantic similarity scores for all possible pairs of 1,417 genes in the Developmental Disorder Gene2Phenotype (DDG2P) database using Human Phenotype Ontology terms. Next, for 62 probands with two molecular diagnoses in the Deciphering Developmental Disorders study, we computed semantic similarity scores between the probands' phenotypes and DDG2P phenotypes associated with the two disorders and compared the results with manual attribution of proband phenotypes to none, one, or both of the genes. We found a spectrum of phenotypic similarity for dual diagnoses, both across all DDG2P genes and within dual diagnosed probands, from phenotypically distinct through blended to indistinguishable conditions. Pairwise semantic similarity scores between two DDG2P genes were a good predictor of the extent of phenotypic blending observed in probands. Dual diagnoses involving genes linked with synergistic phenotypes can result in more extreme presentations while those involving antagonistic phenotypes have spuriously high pairwise semantic similarity scores despite a potentially milder atypical presentation. We suggest that the phenotypic contribution of two molecular diagnoses may contain discrete, synergistic, or antagonistic elements. Conceptual recognition of this phenotypic spectrum is important for making a final clinico-molecular diagnosis and providing accurate genetic counseling.

A Corpus-based Behavioral Profile Study of Synonymous Expressions Meaning “to Start”

Synonymy is an extremely significant component in language learning, and the usage of synonymy is rather flexible and difficult to distinguish. In this research, the author employs a corpus-based behavioral profile approach to study the usage of a set of English synonymous expressions meaning “to start”, namely, start, begin, commence, launch, set out. The author first searches in American Contemporary English Corpus to explore the overall usage frequency and specific genre of each synonymous expression, then randomly selects 500 sentences, and makes manual annotation with 12 ID tag and 46 ID tag levels. Later, hierarchical clustering analysis and correspondence analysis have been employed to examine the semantic differences and usage structures of this set of expressions.The results shows that: ① The overall usage frequency varies greatly—from the highest to the lowest are start, begin, launch, set out, commence. ② They are used in various genres. Start is mainly used in informal styles including blogs, while the other four are mainly used in formal styles, including fictions and magazines. ③ Begin and set out have the greatest semantic similarity; start, begin, and set out form one semantic branch, and commence and launch form another, and there is a huge difference between these two branches. ④ There are great differences between the five expressions and their morphological, syntactic and semantic changes.Quantitative analysis method has been used to reveal the different semantic features and usage of a set of synonymous expressions, which is conducive to both promote the study of synonymous relations, and verify the effectiveness and applicability of the behavioral profile approach in semantic study.

The impact of threat of shock on memory for threat-irrelevant information: Evidence for a role of reduced semantic organization

Research into the effects of acute anxiety on episodic memory has produced inconsistent findings, particularly for threat-neutral information. In two experiments, we tested the hypothesis that anxiety induced by threat of shock can interfere with the use of semantic-organizational processes that benefit memory. In Experiment 1, participants viewed and freely recalled two lists of semantically unrelated neutral words, one encoded in a threatening context (threat blocks) and one encoded without threat (safe blocks). As predicted, significantly fewer words were recalled during threat than safe blocks. Moreover, free recall patterns following threat blocks showed lower levels of semantic organization as assessed using a “path length” measurement that considers the semantic distance between pairs of consecutively recalled words. Both effects unexpectedly interacted with block order, such that they primarily reflected improved recall and increased semantic organization from the first to the second block in participants who received the threat block first. Experiment 2 used semantically related word lists to reduce potential impacts of task experience on semantic organization. Free recall was again less accurate and showed longer (less organized) path lengths for threat than safe blocks, and the path length effect no longer interacted with block order. Moreover, threat-induced changes to path lengths emerged as a mediator of the relation between physiological effects of threat (increased skin conductance) and reduced subsequent memory. These data point to semantic control processes as an understudied determinant of when and how acute anxiety impacts episodic memory.

Extracting Features from Text Flows based on Semantic Similarity for Text Classification: an Approach Inspired by Audio Analysis

Text classification is a mainly investigated challenge in Natural Language Processing (NLP) research. The higher performance of a classification model depends on a representation that can extract valuable information about the texts. Aiming not to lose crucial local text information, a way to represent texts is through flows, sequences of information collected from texts. This paper proposes an approach that combines various techniques to represent texts: the representation by flows, the benefit of the word embeddings text representation associated with lexicon information via semantic similarity distances, and the extraction of features inspired by well-established audio analysis features.In order to perform text classification, this approach splits the text into sentences and calculates a semantic similarity metric to a lexicon on an embedding vector space. The sequence of semantic similarity metrics composes the text flow. Then, the method performs the extraction of twenty-five features inspired by audio analysis (named Audio-Like Features). The features adaptation from audio analysis comes from a similitude between a text flow and a digital signal, in addition to the existing relationship between text, speech, and audio. We evaluated the method in three NLP classification tasks: Fake News Detection in English, Fake News Detection in Portuguese, and Newspaper Columns versus News Classification. The approach efficacy is compared to baselines that embed semantics in text representation: the Paragraph Vector and the BERT. The objective of the experiments was to investigate if the proposed approach could compete with the baselines methods improve their efficacy when associated with them. The experimental evaluation demonstrates that the association between the proposed and the baseline methods can enhance the baseline classification efficacy in all three scenarios. In the Fake News Detection in Portuguese task, our approach surpassed the baselines and obtained the best effectiveness (PR-AUC = 0.98).

Enhanced Prototypical Network with Customized Region-Aware Convolution for Few-Shot SAR ATR

With the prosperous development and successful application of deep learning technologies in the field of remote sensing, numerous deep-learning-based methods have emerged for synthetic aperture radar (SAR) automatic target recognition (ATR) tasks over the past few years. Generally, most deep-learning-based methods can achieve outstanding recognition performance on the condition that an abundance of labeled samples are available to train the model. However, in real application scenarios, it is difficult and costly to acquire and to annotate abundant SAR images due to the imaging mechanism of SAR, which poses a big challenge to existing SAR ATR methods. Therefore, SAR target recognition in the situation of few-shot, where only a scarce few labeled samples are available, is a fundamental problem that needs to be solved. In this paper, a new method named enhanced prototypical network with customized region-aware convolution (CRCEPN) is put forward to specially tackle the few-shot SAR ATR tasks. To be specific, a feature-extraction network based on a customized and region-aware convolution is first developed. This network can adaptively adjust convolutional kernels and their receptive fields according to each SAR image’s own characteristics as well as the semantical similarity among spatial regions, thus augmenting its capability to extract more informative and discriminative features. To achieve accurate and robust target identity prediction under the few-shot condition, an enhanced prototypical network is proposed. This network can improve the representation ability of the class prototype by properly making use of training and test samples together, thus effectively raising the classification accuracy. Meanwhile, a new hybrid loss is designed to learn a feature space with both inter-class separability and intra-class tightness as much as possible, which can further upgrade the recognition performance of the proposed method. Experiments performed on the moving and stationary target acquisition and recognition (MSTAR) dataset, the OpenSARShip dataset, and the SAMPLE+ dataset demonstrate that the proposed method is competitive with some state-of-the-art methods for few-shot SAR ATR tasks.

Enhancing chatbot performance for imaging recommendations: Leveraging GPT-4 and context-awareness for trustworthy clinical guidance

PurposeTo investigate if GPT-4 improves the accuracy, consistency, and trustworthiness of a context-aware chatbot to provide personalized imaging recommendations from American College of Radiology (ACR) appropriateness criteria documents using semantic similarity processing: In addition, we sought to enable auditability of the output by revealing the information source the decision relies on. Material and MethodsWe refined an existing chatbot that incorporated specialized knowledge of the ACR guidelines by upgrading GPT-3.5-Turbo to its successor GPT-4 by OpenAI, using the latest version of LlamaIndex, and improving the prompting strategy. This chatbot was compared to the previous version, generic GPT-3.5-Turbo and GPT-4, and general radiologists regarding the performance in applying the ACR appropriateness guidelines. ResultsThe refined context-aware chatbot performed superior to the previous version using GPT-3.5-Turbo, generic chatbots GPT-3.5-Turbo and GPT-4, and general radiologists in providing “usually or may be appropriate” recommendations according to the ACR guidelines (all p < 0.001). It also outperformed GPT-3.5-Turbo and general radiologists in respect to “usually appropriate” recommendations (both p < 0.001). Moreover, the consistency in correct answers was higher with 78 % consistent correct “usually appropriate” answers and 94 % for “usually or may be appropriate” recommendations. In all cases, the same source documents were chosen, ensuring transparency. ConclusionOur study demonstrates the significance of context awareness in ensuring the use of appropriate knowledge and proposes a strategy to enhance trust in chatbot-based outputs to provide transparency. The improvements in accuracy, consistency, and source transparency address trust issues and enhance the clinical decision support process.Abbreviations: ACR, American College of Radiology; accGPT, appropriateness criteria context aware GPT; accGPT-4, appropriateness criteria context aware GPT using GPT-4; GPT, generative pre-trained transformer; LLM, Large Language Model.

High School Students’ Perceptions of Nursing as a Future Career Choice: A Qualitative Study

Background: Understanding the perceptions of high school students, before entering the field of nursing, can be a basis for future planning in this regard. Accordingly, this study aimed to explain the perceptions of these students about nursing as a future career choice in Iran. Methods: This qualitative content analysis study was conducted in four public high schools in Tehran, Iran. Participants included 28 students recruited using purposive sampling who were administered semi-structured, face-to-face interviews in 2019. The duration of the interviews ranged from 30 to 45 minutes and the interviews were transcribed verbatim. The method proposed by Elo and Kyngäs was used to analyze the data. Initially, semantic similarities and differences were reviewed and the subcategories were identified. After repeated reviews, the related subcategories were placed in one category. Results: Data analysis showed students’ perceptions of nursing as a future career could be classified into three categories and seven subcategories. The main categories included a positive view of the nursing profession (with two subcategories: interest in the profession and being a supporter), barriers to choosing nursing as a future career (with three subcategories: the lack of professional independence, poor financial support, and negative evaluation of nursing), and cultural issues (with two subcategories: stigmatized public image and nursing as a gender-dominated profession). Conclusion: The results of this study showed high school students have positive and negative views towards nursing and the influence of cultural issues on the nursing profession. If those interested in nursing can enter this field, they will act as professionals who provide optimal patient care.

The Sharing Model of Kindergarten Art Enlightenment Education Curriculum Resources Based on Deep Learning Algorithm

In order to meet the autonomous needs of children’s art learning, this paper studies the sharing model of kindergarten art enlightenment education curriculum resources under the deep learning algorithm. Build a kindergarten art enlightenment education curriculum resource-sharing model that includes the classification and storage layer of education curriculum resources, the metadata annotation layer of education curriculum resources and the shared business service layer. The deep learning algorithm is used in the classification and storage layer of education curriculum resources. The education curriculum resources are divided into courseware resources, case resources and media materials resources through unsupervised restricted Boltzmann machine (RBM) coding and supervised training process. The ontology model of various educational curriculum resources is constructed, and the semantic information of educational curriculum resources is described using ontology. The metadata annotation layer of education curriculum resources annotates the metadata semantics of education curriculum resources through the constructed ontology, which improves the consistency and sharing of education curriculum resources management. The implementation of resource sharing business service is based on semantic analysis and query of educational curriculum resources. Therefore, the shared business service layer adopts a semantic query method based on improved similarity algorithm, which combines semantic distance, node density, node depth and relationship type to build a semantic similarity calculation method to achieve semantic analysis and query of educational curriculum resources. The experimental results show that the model is suitable for multi-classification of shared resources in the kindergarten art enlightenment education curriculum, and the semantic query accuracy is high in the process of sharing. When the threshold is [Formula: see text] 0.5, the query effect of the sharing model in this paper is ideal, which can meet the needs of users’ autonomous learning.