Text Matching Model Research Articles

Image–Text Matching Model Based on CLIP Bimodal Encoding

Image–text matching is a fundamental task in the multimodal research field, connecting computer vision and natural language processing by aligning visual content with corresponding textual descriptions. Accurate matching is critical for applications such as image captioning and text-based image retrieval yet remains challenging due to the differences in data modalities. This paper addresses these challenges by proposing a robust image–text matching model inspired by Contrastive Language–Image Pre-training (CLIP). Our approach employs the Vision Transformer (ViT) model as the image encoder and Bidirectional Encoder Representations from Transformers (Bert) as the text encoder, integrating these into a shared vector space to measure semantic similarity. We enhance the model’s training efficiency using the LiT-tuning paradigm to optimize learning through a cosine decay strategy for dynamic adjustment of the learning rate. We validate our method on two benchmark datasets, WuKong and Flickr30k, demonstrating that our model achieves superior performance and significantly improves key evaluation metrics. The results underscore the model’s effectiveness in achieving accurate and robust image–text alignment.

Exploring highly concise and accurate text matching model with tiny weights

Exploring highly concise and accurate text matching model with tiny weights

A Novel Address-Matching Framework Based on Region Proposal

Geocoding is a fundamental component of geographic information science that plays a crucial role in various geographical studies and applications involving text data. Current mainstream geocoding methods fall into two categories: geodesic-grid prediction and address matching. However, the geodesic-grid-prediction method’s localization accuracy is hindered by the density of grid partitioning, struggling to strike a balance between prediction accuracy and grid density. Address-matching methods mainly focus on the semantics of query text. However, they tend to ignore keyword information that can be used to distinguish candidates and introduce potential interference, which reduces matching accuracy. Inspired by the human map-usage process, we propose a two-stage address-matching approach that integrates geodesic-grid prediction and text-matching models. Initially, a multi-level text-classification model is used to generate a retrieval region proposal for an input query text. Subsequently, we search for the most relevant point of interest (POI) within the region-proposal area using a semantics-based text-retrieval model. We evaluated the proposed method using POI data from the Beijing Chaoyang District. The experimental results indicate that the proposed method provides high address-matching accuracy, increasing Recall@1 by 0.55 to 1.56 percentage points and MRR@5 by 0.54 to 1.68 percentage points.

IMAI: A short text matching model integrating multi-angle information

The short text matching models can be roughly divided into representation-based and interaction-based approaches. However, current representation-based text matching models often lack the ability to handle sentence pairs and typically only perform feature interactions at the network’s top layer, which can lead to a loss of semantic focus. The interactive text matching model has significant shortcomings in extracting differential information between sentences and may ignore global information. To address these issues, this article proposes a model structure that combines a dual-tower architecture with an interactive component, which compensates for their respective weaknesses in extracting sentence semantic information. Simultaneously, a method for integrating semantic information is proposed, enabling the model to capture both the interactive information between sentence pairs and the differential information between sentences, thereby addressing the issues with the aforementioned approaches. In the process of network training, a combination of cross-entropy and cosine similarity is used to calculate the model loss. The model is optimized to a stable state. Experiments on the commonly used datasets of QQP and MRPC validate the effectiveness of the proposed model, and its performance is stably improved.

Seeing both sides: context-aware heterogeneous graph matching networks for extracting-related arguments

Our research focuses on extracting exchanged views from dialogical documents through argument pair extraction (APE). The objective of this process is to facilitate comprehension of complex argumentative discourse by finding the related arguments. The APE comprises two stages: argument mining and argument matching. Researchers typically employ sequence labeling models for mining arguments and text matching models to calculate the relationships between them, thereby generating argument pairs. However, these approaches fail to capture long-distance contextual information and struggle to fully comprehend the complex structure of arguments. In our work, we propose the context-aware heterogeneous graph matching (HGMN) model for the APE task. First, we design a graph schema specifically tailored to argumentative texts, along with a heterogeneous graph attention network that effectively captures context information and structural information of arguments. Moreover, the text matching between arguments is converted into a graph matching paradigm and a multi-granularity graph matching model is proposed to handle the intricate relationships between arguments at various levels of granularity. In this way, the semantics of argument are modeled structurally and thus capture the complicated correlations between arguments. Extensive experiments are conducted to evaluate the HGMN model, including comparisons with existing methods and the GPT series of large language models (LLM). The results demonstrate that HGMN outperforms the state-of-the-art method.

STMAP: A novel semantic text matching model augmented with embedding perturbations

Semantic text matching models have achieved outstanding performance, but traditional methods may not solve Few-shot learning problems and data augmentation techniques could suffer from semantic deviation. To solve this problem, we propose STMAP, which is implemented from the perspective of data augmentation based on Gaussian noise and Noise Mask signal. We also employ an adaptive optimization network to dynamically optimize the several training targets generated by data augmentation. We evaluated our model on four English datasets: MRPC, SciTail, SICK, and RTE, with achieved scores of 90.3%, 94.2%, 88.9%, and 68.8%, respectively. Our model obtained state-of-the-art (SOTA) results on three of the English datasets. Furthermore, we assessed our approach on three Chinese datasets, and achieved an average improvement of 1.3% over the baseline model. Additionally, in the Few-shot learning experiment, our model outperformed the baseline performance by 5%, especially when the data volume was reduced by around 0.4. Our ablation experiments further validated the effectiveness of STMAP.22We have released our source code. https://github.com/wangyanhao0517/STMAP.

Mapping Relationship Discovery of Multidimensional Architectures in Autonomous Transportation System Based on Text-Matching Model

Mapping relationships of multidimensional architectures play an essential role in the autonomous transportation system (ATS), which can help understand the complex relationships between multidimensional architectures. The current mapping relationship discovery for multidimensional architectures in ATS requires significant manual involvement, leading to the underutilization of textual data and the intense subjectivity of results. In order to address the above issues, it is necessary to mine and further utilize the semantic information in the textual data. This study applies the text-matching model to the mapping relationship discovery of multi-dimensional architectures, which can calculate the semantic similarity between texts. On this basis, a method based on the Siamese-BERT-wwm-Bi-LSTM model is proposed, which incorporates Chinese BERT with whole word masking (BERT-wwm), bidirectional long-short term memory (Bi-LSTM), and the Siamese Network. A series of experiments are conducted with different text-matching models. The results show that the precision rate, recall rate, and F1-score exceed 80% for most applied methods, which verifies the feasibility of using the text-matching model for mapping relationship discovery. These results expect to provide a method with good performance that can automatically perform mapping relationship discovery.

An EHR Data Quality Evaluation Approach Based on Medical Knowledge and Text Matching

Introduction: Recently, medical artificial intelligence based on Electronic Health Records (EHR) is a significant research field, and EHR data has been widely used in clinical decision support systems and medical diagnosis systems. However, because EHR are used to record the patient's disease information and are not primarily designed for research and discovery, the utility of EHR for research will be hindered by data quality problems. Therefore, it is a meaningful and challenging task to evaluate the data quality of EHR before they are used in medical artificial intelligence. Most of the current EHR data quality evaluation methods are based on some conventional evaluation indicators, and rarely consider the introduction of clinical evidence.Materials and methods: we propose an EHR data quality evaluation approach based on clinical evidence and a deep text matching model. First, based on the medical knowledge of the particular disease, we establish the list of standard clinical evidence descriptions including typical symptoms and special signs, etc. Then we find the relevant clinical evidence from the EHR based on the text matching model, and finally evaluate the quality of the EHR based on the quantity and quality of the relevant clinical evidence found.Results: The experimental results of more than 1,000 EHR for two types of diseases show that our approach can effectively distinguish high-quality EHR from low-quality EHR, and the high-quality EHR found generally contains sufficient and consistent information related to disease diagnosis.Conclusions: Experiments results on a real-world dataset demonstrate the effectiveness of our EHR data quality evaluation approach based on medical knowledge and text matching.

Text Matching in Insurance Question-Answering Community Based on an Integrated BiLSTM-TextCNN Model Fusing Multi-Feature

Along with the explosion of ChatGPT, the artificial intelligence question-answering system has been pushed to a climax. Intelligent question-answering enables computers to simulate people's behavior habits of understanding a corpus through machine learning, so as to answer questions in professional fields. How to obtain more accurate answers to personalized questions in professional fields is the core content of intelligent question-answering research. As one of the key technologies of intelligent question-answering, the accuracy of text matching is related to the development of the intelligent question-answering community. Aiming to solve the problem of polysemy of text, the Enhanced Representation through Knowledge Integration (ERNIE) model is used to obtain the word vector representation of text, which makes up for the lack of prior knowledge in the traditional word vector representation model. Additionally, there are also problems of homophones and polyphones in Chinese, so this paper introduces the phonetic character sequence of the text to distinguish them. In addition, aiming at the problem that there are many proper nouns in the insurance field that are difficult to identify, after conventional part-of-speech tagging, proper nouns are distinguished by especially defining their parts of speech. After the above three types of text-based semantic feature extensions, this paper also uses the Bi-directional Long Short-Term Memory (BiLSTM) and TextCNN models to extract the global features and local features of the text, respectively. It can obtain the feature representation of the text more comprehensively. Thus, the text matching model integrating BiLSTM and TextCNN fusing Multi-Feature (namely MFBT) is proposed for the insurance question-answering community. The MFBT model aims to solve the problems that affect the answer selection in the insurance question-answering community, such as proper nouns, nonstandard sentences and sparse features. Taking the question-and-answer data of the insurance library as the sample, the MFBT text-matching model is compared and evaluated with other models. The experimental results show that the MFBT text-matching model has higher evaluation index values, including accuracy, recall and F1, than other models. The model trained by historical search data can better help users in the insurance question-and-answer community obtain the answers they need and improve their satisfaction.

Improved Text Matching Model Based on BERT

Text matching is a basic and important task in natural language understanding, this paper proposes a new model BBMC for the problem of insufficient feature extraction ability of existing text matching models, which integrates BiLSTM and multi-scale CNN on the basis of BERT. First, the word embedding representation of the text is obtained by the BERT, and then the semantic features of the text are further extracted by the double-layer BiLSTM, followed by the multi-scale CNN model, the key local features are extracted, and finally the linear and SoftMax function are used to classify. Experimental results on the LCQMC dataset show that the BBMC has been improved to a certain extent compared with other methods, and the accuracy on the test set can be best achieved 88.01%.

Lightweight Text Matching Method

This paper constructs a lightweight text matching model. This model fully extracts the features of two text through the fusion of double tower and interactive methods, and then conducts deep interaction. Then, it classifies the last extracted text features through the classification network, and conducts training, validation and testing on the Chinese text matching dataset LCQMC. Although the experimental results show that the model in this paper is relatively good, it still needs to be improved.

A text matching model based on dynamic multi‐mask and augmented adversarial

AbstractThe text matching is a basic task of NLP and is important for tasks such as text retrieval, question answering, and so forth. The development of pre‐trained language models has promoted the progress of text matching tasks. However, due to the natural particularity of Chinese characters and expressions, the Chinese text matching tasks still have problems such as word segmentation difficulty, serious semantic loss, and model instability. In this paper, we propose the DAINet model, which includes DMM, AA and IO modules. We use the Dynamic Multi‐Mask module (DMM) to enhance the completeness of word segmentation. Then we use the Augmented Adversarial module (AA) to further extraction of semantic information. Finally, we use the Integrated Output module (IO) for a more stable output. We conducted experiments on LCQMC, BQ and Xiaobu datasets and compared the results with seven strong baseline models. The results showed that DAINet model made great improvement, including improving ACC value of BQ dataset to , AUC value to , ACC value of LCQMC dataset to and AUC value to . The ACC value of Xiaobu dataset was improved to and the AUC value was improved to . Further ablation experiment results show that the proposed DMM, AA and IO modules have good adaptability and improvement over existing models.

Question and Answer Techniques for Financial Audits in Universities Based on Deep Learning

Financial auditing in universities is highly specialized, with a huge knowledge system and rapid updates. Auditors will encounter various problems and situations in their work and need to acquire domain knowledge efficiently and accurately to solve the difficulties they encounter. The existing audit information software, however, is mostly aimed at the management of audit affairs and lacks the relevant functions to acquire and retrieve knowledge of specific audit domains. In this study, we use deep learning theory as support to conduct an in-depth study on the key technologies of question and answer systems in the field of financial auditing in universities. In the question-answer retrieval stage, the local information and the global information of the sentence are first modelled using a two-way coding model based on the attentional mechanism, and then, an interactive text matching model is used to interact directly at the input layer, and a multilayer convolutional neural network model cable news network (CNN) is used to extract the fine-grained matching features from the interaction matrix; this study adopts two matching methods. We have conducted comparative experiments to verify the effectiveness and application value of the entity recognition algorithm based on this study’s algorithm and the question-answer retrieval model based on multi-granularity text matching in the university financial audit domain.

Short text matching model with multiway semantic interaction based on multi-granularity semantic embedding

Short text matching model with multiway semantic interaction based on multi-granularity semantic embedding

Design and Analysis of Network Virus Defense System Based on Multimodal Data Mining Technology

This study analyzes and designs a network security management system based on the K-means algorithm. Through the investigation of network security managers, this article uses structured modeling technology to derive the system’s logical business functions, which are data acquisition function, data analysis function, and post-processing function. This study uses a three-tier architecture to design the system, describes the system data business processing flow, and gives the key flow of the K-means algorithm application. In addition, the system database is designed to provide support for system data processing. This study proposes a dual-network text matching model based on local interactive components. This model composes two modalities of single-modal short text data: a positional structural modal for describing local interactions and a global semantic understanding modal for extracting global semantic information. Two heterogeneous networks are used to extract two modalities. The principle of complementarity of modalities is realized by constructing the differences of each modal. At the same time, through the attention mechanism, the position information of the position structure modal is transferred to the global semantic understanding modal to obtain consistent comprehensive information so as to realize the principle of modal consistency. On this basis, by designing low-order interaction functions and high-order interaction functions, and using long- and short-term memory networks, we have, respectively, improved the position structure modal and global semantic understanding modal. The theoretical analysis and simulation results of the model show that the propagation characteristics of the network worm are completely determined by the threshold R0 required for its existence. When R0 >1, the network worm will become popular even with a small initial infection number; conversely, even if the initial infection machine is large, the network worm will eventually become extinct. The research results of this article also show that the introduction of mobile devices has an important impact on the defense technology of network worms. To curb network worms that use both the network and mobile devices to spread, and to increase the proportion of machines that have never been infected with the virus, the most effective method is to control the number of mobile devices and reduce the possibility of cross-infection between mobile devices and machines. The simulation results show that, regardless of whether dynamic isolation and antivirus software are considered, the local area network-based choke method given in this study can effectively curb intelligent network worms that have slow scanning rates and clear scanning targets. In addition, the choke method presented in this article can determine the choke threshold without affecting the normal network scanning behavior of users in the local area network.

Product packaging text and composition colour matching model design based on visual communication

In order to overcome the problems of low pixel extraction accuracy and long matching time in traditional models, a product packaging text and composition colour matching model based on visual communication was designed. Under the concept of visual communication, the phased feature matching method is used to extract the text and composition pixels in the product packaging in blocks. The morphological transformation method is used to construct the regional pixel threshold denoising function, and the colour representation model of text and composition is constructed. The adaptive block feature matching method is used to match the colour of text pixels and composition pixels to achieve the matching of product packaging text and composition colours. The experimental results show that the designed model has high pixel extraction accuracy, and the shortest time is only 2.5 s, which fully shows that the designed model has strong practical application performance.

Product packaging text and composition colour matching model design based on visual communication

Product packaging text and composition colour matching model design based on visual communication

Pseudo-siamese networks with lexicon for Chinese short text matching

Short text matching is one of the fundamental technologies in natural language processing. In previous studies, most of the text matching networks are initially designed for English text. The common approach to applying them to Chinese is segmenting each sentence into words, and then taking these words as input. However, this method often results in word segmentation errors. Chinese short text matching faces the challenges of constructing effective features and understanding the semantic relationship between two sentences. In this work, we propose a novel lexicon-based pseudo-siamese model (CL2 N), which can fully mine the information expressed in Chinese text. Instead of utilizing a character-sequence or a single word-sequence, CL2 N augments the text representation with multi-granularity information in characters and lexicons. Additionally, it integrates sentence-level features through single-sentence features as well as interactive features. Experimental studies on two Chinese text matching datasets show that our model has better performance than the state-of-the-art short text matching models, and the proposed method can solve the error propagation problem of Chinese word segmentation. Particularly, the incorporation of single-sentence features and interactive features allows the network to capture the contextual semantics and co-attentive lexical information, which contributes to our best result.

A simple and efficient text matching model based on deep interaction

In recent years, text matching has gained increasing research focus and shown great improvements. However, due to the long-distance dependency and polysemy, existing text matching models cannot effectively capture the contextual and implicit semantic information of texts. Additionally, existing models are lack of generalization ability when applied to different scenarios. In this study, we propose a novel Deep Interactive Text Matching (DITM) model by integrating the encoder layer, the co-attention layer, and the fusion layer as an interaction module, based on a matching-aggregation framework. In particular, the interaction process is iterated multiple times to obtain the in-depth interaction information, and the relationship between the text pair is extracted through the multi-perspective pooling. We conduct extensive experiments on four text matching tasks, i.e., opinion retrieval, answer selection, paraphrase identification and natural language inference. Compared with the state-of-the-art text matching methods, the proposed model achieved the best results on most of the tasks, which proves that our model could effectively capture the interactive information between text pairs, and has a high generalization ability among different tasks. Further multi-lingual investigations show the similarities of the performance between English and Chinese, which suggest that our model could be ported to other languages. The research contributes a simple and efficient implementation of text matching in a situation where there is limited computing capacity, and sheds light on leveraging text matching models to facilitate a range of downstream tasks.

Hierarchical matching network for multi-turn response selection in retrieval-based chatbots

Proper response selection is a crucial challenge in retrieval-based chatbots. The state-of-the-art methods match a response with the word sequence of a context, or match the response with each utterance in the context and then accumulate matching information. The former architecture could lose some important local matching information in utterance–response pairs and does not explicitly capture the relationships and dependencies among utterances. The latter architecture does not consider the important global matching information because there is no match between the response and the context at word level. Hence, the above methods have a problem, without considering the fact that matching a response with different levels of a context could match different information for multi-turn response selection. In this work, we propose a hierarchical matching network to match a response with the word and utterance level of a context. At word level, we concatenate the multi-turn context as a long word sequence and then adopt a text matching model to match the response with the word sequence which can capture important matching information at word level. At utterance level, we employ the identical text matching model to match the response with each utterance in the context to capture important matching information for each utterance–response pair and then accumulate the matching information by a recurrent neural network to model the relationships of utterances. At last, the hierarchical matching information is fused to get the final matching information. Experiments on two large-scale public multi-turn response selection datasets show that the proposed model significantly outperforms the state-of-the-art baseline models.