Search

With the advancement of technology in recent years, the application of artificial intelligence in real life has become more extensive. Graphic recognition is a hot spot in the current research of related technologies. It involves machines extracting key information from pictures and combining it with natural language processing for in-depth understanding. Existing methods still have obvious deficiencies in fine-grained recognition and deep understanding of contextual context. Addressing these issues to achieve high-quality image-text recognition is crucial for various application scenarios, such as accessibility technologies, content creation, and virtual assistants. To tackle this challenge, a novel approach is proposed that combines the Mask R-CNN, DCGAN, and ALBERT models. Specifically, the Mask R-CNN specializes in high-precision image recognition and segmentation, the DCGAN captures and generates nuanced features from images, and the ALBERT model is responsible for deep natural language processing and semantic understanding of this visual information. Experimental results clearly validate the superiority of this method. Compared to traditional image-text recognition techniques, the recognition accuracy is improved from 85.3% to 92.5%, and performance in contextual and situational understanding is enhanced. The advancement of this technology has far-reaching implications for research in machine vision and natural language processing and open new possibilities for practical applications.

In natural language processing (NLP), a Question Answering System (QAS) refers to a system or model that is designed to understand and respond to user queries in natural language. As we navigate through the recent advancements in QAS, it can be observed that there is a paradigm shift of the methods used from traditional machine learning and deep learning approaches towards transformer-based language models. While significant progress has been made, the utilization of these models for historical QAS and the development of QAS for Malay language remain largely unexplored. This research aims to bridge the gaps, focusing on developing a Multilingual QAS for history of Malaysia by utilizing a transformer-based language model. The system development process encompasses various stages, including data collection, knowledge representation, data loading and pre-processing, document indexing and storing, and the establishment of a querying pipeline with the retriever and reader. A dataset with a collection of 100 articles, including web blogs related to the history of Malaysia, has been constructed, serving as the knowledge base for the proposed QAS. A significant aspect of this research is the use of the translated dataset in English instead of the raw dataset in Malay. This decision was made to leverage the effectiveness of well-established retriever and reader models that were trained on English data. Moreover, an evaluation dataset comprising 100 question-answer pairs has been created to evaluate the performance of the models. A comparative analysis of six different transformer-based language models, namely DeBERTaV3, BERT, ALBERT, ELECTRA, MiniLM, and RoBERTa, has been conducted, where the effectiveness of the models was examined through a series of experiments to determine the best reader model for the proposed QAS. The experimental results reveal that the proposed QAS achieved the best performance when employing RoBERTa as the reader model. Finally, the proposed QAS was deployed on Discord and equipped with multilingual support through the incorporation of language detection and translation modules, enabling it to handle queries in both Malay and English. Doi: 10.28991/ESJ-2024-08-02-019 Full Text: PDF

Spam emails, which are unsolicited bulk emails, pose a significant threat in digital communication security. To counter spam emails, learning-based spam email filtering methods have been extensively studied. However, as spam patterns evolve, these methods face challenges in maintaining the accuracy of models trained on outdated patterns. To demonstrate these limitations empirically and gain insight into the classification patterns of spam email filtering models, we propose an advanced analysis method to analyze the performance degradation of spam email filtering models. The proposed analysis method involves text preprocessing, embedding model training, spam email filtering model training, evaluation, and analysis of the classification patterns of the learning-based spam email filtering models. From the experimental results under various datasets and spam email filtering models, we show that the accuracy of spam email filtering models significantly decreases when the feature distribution of the test dataset is different from the training dataset. We also provides valuable insights for improving the model architecture, dataset structure, and training strategies by analysis of various factors such as confusion matrix, performance metrics, mean sequence length, out-of-vocabulary (OOV) rate, and top-20 tokens.

As an extension of the transformer architecture, the BERT model has introduced a new paradigm for natural language processing, achieving impressive results in various downstream tasks. However, high-performance BERT-based models—such as ELECTRA, ALBERT, and RoBERTa—suffer from limitations such as poor continuous learning capability and insufficient understanding of domain-specific documents. To address these issues, we propose the use of an attention mechanism to combine BERT-based models with neural topic models. Unlike traditional stochastic topic modeling, neural topic modeling employs artificial neural networks to learn topic representations. Furthermore, neural topic models can be integrated with other neural models and trained to identify latent variables in documents, thereby enabling BERT-based models to sufficiently comprehend the contexts of specific fields. We conducted experiments on three datasets—Movie Review Dataset (MRD), 20Newsgroups, and YELP—to evaluate our model’s performance. Compared to the vanilla model, the proposed model achieved an accuracy improvement of 1–2% for the ALBERT model in multiclassification tasks across all three datasets, while the ELECTRA model showed an accuracy improvement of less than 1%.

Named entity recognition of wheat diseases and pests is a key step to building a knowledge graph. Aiming at the problems of lack of training data, complex entity structure, diverse entity types and uneven entity distribution in wheat diseases and pests field, under the promise of fully mining the implicit knowledge, two data augmentation methods are used to expand sentence semantic information, and to construct the corpus WpdCNER (wheat pests and diseases Chinese named entity recognition) and the field lexicon WpdDict (wheat pests and diseases dictionary). And 16 categories of entities are defined with the field experts’ guidance. Meanwhile, Chinese named entity recognition model based on rules amendment WPD-RA (wheat pests and disease-rules amendment model) is proposed. This model is carried out entity recognition based on ALBERT+BiLSTM+CRF (a lite bi-directional encoder representation from transformer + bi-directional long short-term memory + conditional random field), and specific rules are defined to amend entity boundaries after recognition. The WPD-RA model achieves the best results with 94.72% precision, 95.23% recall, and 94.97% F1. Its precision is increased by 1.71 percentage points, recall is increased by 0.34 percentage points, and F1 is increased by 1.03 percentage points, compared with the model without rules. Experimental results show that the model can effectively recognize named entities in wheat diseases and pests field, and its performance is better than other models. The proposed model provides a reference idea for named entity recognition task in other fields such as food safety and biology.

Information retrieval and text mining fields extensively utilize text semantic matching models. In this paper, civil litigation Q&A under the environmental justice system is taken as a specific research field, and after constructing a civil litigation Q&A system based on deep learning, two of the key techniques—question categorization and semantic matching—are selected as the main research content. Specifically, the ALBERT algorithm is used to extract word vectors, and the hidden feature vectors are obtained through BiLSTM modeling of contextual relationships and then combined with the Attention mechanism for scoring and weighting to obtain the final text-level vectors for classification so as to establish the civil litigation question classification model based on ALBERT. Then, we establish the BERT-based civil litigation question and answer matching model by sorting the set of candidate answers by semantic matching degree based on the BERT algorithm. Selected datasets and comparison algorithms are experimented with, and the analysis shows that the question classification model has a better effect than civil litigation question text classification, and the values of each index have been improved by 0.75%~3.00% on the basis of the baseline model. The MAP and MRR values (0.76~0.86) of the question-matching model are higher than those of the comparison model, verifying its superior performance in semantically assigning characters. The model proposed in this paper is more useful because it can provide civil litigation counseling to the public.

A model is utilized in this paper to analyze the textual emotions of ideology in Marxist theory. Long and short-term memory networks are chosen as the main method of text analysis to construct the main process of Marxist ideology education. Combined with the hybrid self-attention mechanism, the efficiency of extracting data features from the text was improved. The results show that the ALBERT-SABL-based sentiment analysis model is 86.9% accurate in extracting the sentiment of the ideology, and the F1 value is 87.6%. Compared with TextCnn, the accuracy has improved by 1.8%. Different schools have different levels of identification with Marx’s ideology, and under the identification dimension, the highest sentiment identification dimension among the eight school samples is School 2, with a dimension of 100. This study provides reference data in the ideological education of Marxist theory and promotes the development of Marx’s ideology.

The purpose of this essay is threefold. Firstly, it attempts to introduce the diplomatic and collecting careers of the Austro-Hungarian diplomat Baron Albert Eperjesy (1848–1916), who was the highest representative of his country in numerous European capitals and –between 1895 and 1901– Tehran. Secondly, an attempt will be made to contextualise his collecting habits by drawing attention to the peculiarities of Austro-Hungarian collector diplomats. Finally, and perhaps most importantly, the Persian element of this collection will be discussed within the previously outlined framework, namely, what artworks it did include, how and where he obtained them, and what would be their subsequent fate.

The classification performance is highly affected because hyperspectral images include many bands, have high dimensions, and have few labeled training samples. This challenge is reduced by using rich spatial information and an effective classifier. The classifiers in this study are BERT-based (Bidirectional Encoder Representations from Transformers) models, which have recently been applied in natural language processing. The BERT model and its performance-improved version, the ALBERT (A Lite BERT) model, are utilized as transformer-based models. Because of their structure, these models can also accept spatial information via ‘segment embeddings’. Segmentation algorithms are commonly used in the literature to get spatial information. Superpixel methods have shown superior results in the segmentation literature due to the utility of working at the superpixel level rather than the conventional pixel level. HyperSLIC, a modified version of the SLIC superpixel method for hyperspectral images, is employed as input to BERT-based models in this study. In addition, HyperSLIC segmentation results are merged with the DBSCAN algorithm for similar superpixels to increase the size of spatially similar areas and called as HyperSLIC-DBSCAN. The effects of segment embedding information on classification accuracy in BERT-based models is studied experimentally. Experimental results show that BERT-based models outperform conventional and deep learning-based 1D/2D convolutional neural network classifiers when spatial information is used with the help of segment embedding information.

Football is one of the most popular sports in the world, arousing a wide range of research topics related to its off- and on-the-pitch performance. The extraction of football entities from football news helps to construct sports frameworks, integrate sports resources, and timely capture the dynamics of the sports through visual text mining results, including the connections among football players, football clubs, and football competitions, and it is of great convenience to observe and analyze the developmental tendencies of football. Therefore, in this paper, we constructed a 1000,000-word Chinese corpus in the field of football and proposed a BiLSTM-based model for named entity recognition. The ALBERT-BiLSTM combination model of deep learning is used for entity extraction of football textual data. Based on the BiLSTM model, we introduced ALBERT as a pre-training model to extract character and enhance the generalization ability of word embedding vectors. We then compared the results of two different annotation schemes, BIO and BIOE, and two deep learning models, ALBERT-BiLSTM-CRF and ALBERT BiLSTM. It was verified that the BIOE tagging was superior than BIO, and the ALBERT-BiLSTM model was more suitable for football datasets. The precision, recall, and F-Score of the model were 85.4%, 83.47%, and 84.37%, correspondingly.

The rapid development of artificial intelligence (AI) has injected new vitality into various industries and provided new ideas for the development of traditional Chinese medicine (TCM). The combination of AI and TCM provides more technical support for TCM auxiliary diagnosis and treatment. In the history of TCM, many methods of syndrome differentiation have been observed, among which the differentiation of Zang-fu organs is one of the important methods. The purpose of this paper is to provide support for the localization of Zang-fu in TCM through AI technology. Localization of Zang-fu organs is a method of determining the location of lesions in such organs and is an important stage in the differentiation of Zang-fu organs in TCM. In this paper, the localization model of TCM Zang-fu organs through the neural network model was established. Through the input of symptom text information, the corresponding Zang-fu label for a lesion could be output to provide support for the realization of Zang-fu syndrome differentiation in TCM-assisted diagnosis and treatment. In this paper, the localization of Zang-fu organs was abstracted as multi-label text classification in natural language processing. Using the medical record data of TCM, a Zang-fu localization model based on pretraining models a lite BERT (ALBERT) and bidirectional gated recurrent unit (Bi-GRU) was proposed. Comparison and ablation experiments finally show that the proposed method is more accurate than multilayer perceptron and the decision tree. Moreover, using an ALBERT pretraining model for text representation effectively improves the accuracy of the localization model. In terms of model parameters, the ALBERT pretraining model greatly reduces the number of model parameters compared with the BERT model and effectively reduces the model size. Finally, the F1-value of the Zang-fu localization model proposed in this paper reaches 0.8013 on the test set, which provided certain support for the TCM auxiliary diagnosis and treatment.

Aspect-level sentiment classification aims to solve the problem, which is to judge the sentiment tendency of each aspect in a sentence with multiple aspects. Previous works mainly employed Long Short-Term Memory (LSTM) and Attention mechanisms to fuse information between aspects and sentences, or to improve large language models such as BERT to adapt aspect-level sentiment classification tasks. The former methods either did not integrate the interactive information of related aspects and sentences, or ignored the feature extraction of sentences. This paper proposes a novel multi-grained attention representation with ALBERT (MGAR-ALBERT). It can learn the representation that contains the relevant information of the sentence and the aspect, while integrating it into the process of sentence modeling with multi granularity, and finally get a comprehensive sentence representation. In Masked LM (MLM) task, in order to avoid the influence of aspect words being masked in the initial stage of the pre-training, the noise linear cosine decay is introduced into $n-gram$ . We implemented a series of comparative experiments to verify the effectiveness of the method. The experimental results show that our model can achieve excellent results on Restaurant dataset with numerous number of parameters reduced, and it is not inferior to other models on Laptop dataset.

In order to solve the problem that most aspect level sentiment analysis networks cannot extract the global and local information of the context at the same time. This study proposes an aspect level sentiment analysis model named Combining with A Lite Bidirection Encoder Represention from TransConvs and ConvNets(ALBERTC-CNN). First, the global sentence information and local emotion information in a text are extracted by the improved ALBERTC network, and the input aspect level text is represented by a word vector. Then, the feature vector is mapped to the emotion classification number by a linear function and a softmax function. Finally, the aspect level sentiment analysis results are obtained. The proposed model is tested on two datasets of the SemEval-2014 open task, the laptop and restaurant datasets, and compared with the traditional networks. The results show that compared with the traditional network, the classification accuracy of the proposed model is improved by approximately 4% and 5% on the two sets, whereas the F1 value is improved by approximately 4% and 8%. Additionally, compared with the original ALBERT network, the accuracy is improved by approximately 2%, and the F1 value is improved by approximately 1%.

Microblog is an important platform for mining public opinion, and it is of great value to conduct emotional analysis of microblog texts during the current epidemic. Aiming at the problem that most current emotional classification methods cannot effectively extract deep text features, and that traditional word vectors cannot dynamically obtain the semantics of words according to their context, which leads to classification bias, this research put forward a microblog text emotion classification algorithm based on TCN-BiGRU and dual attention (TCN-BiGRU-DATT). First, the vector representation of the text was obtained using ALBERT. Second, the TCN and BiGRU networks were used to extract the emotional information contained in the text through dual pathway feature extraction, to efficiently obtain the deep semantic features of the text. Then, the dual attention mechanism was introduced to allocate the global weight of the key information in the semantic features, and the emotional features were spliced and fused. Finally, the Softmax classifier was applied for emotion classification. The findings of a comparative experiment on a set of microblog text comments collected throughout the pandemic revealed that the accuracy, recall, and F1 value of the emotion classification method proposed in this paper reached 92.33%, 91.78%, and 91.52%, respectively, which was a significant improvement compared with other models.

Named entity recognition can deeply explore semantic features and enhance the ability of vector representation of text data. This paper proposes a named entity recognition method based on multi-head attention to aim at the problem of fuzzy lexical boundary in Chinese named entity recognition. Firstly, Word2vec is used to extract word vectors, HMM is used to extract boundary vectors, ALBERT is used to extract character vectors, the Feedforward-attention mechanism is used to fuse the three vectors, and then the fused vectors representation is used to remove features by BiLSTM. Then multi-head attention is used to mine the potential word information in the text features. Finally, the text label classification results are output after the conditional random field screening. Through the verification of WeiboNER, MSRA, and CLUENER2020 datasets, the results show that the proposed algorithm can effectively improve the performance of named entity recognition.

Intent classification and named entity recognition of medical questions are two key subtasks of the natural language understanding module in the question answering system. Most existing methods usually treat medical queries intent classification and named entity recognition as two separate tasks, ignoring the close relationship between the two tasks. In order to optimize the effect of medical queries intent classification and named entity recognition tasks, a multi-task learning model based on ALBERT-BILSTM is proposed for intent classification and named entity recognition of Chinese online medical questions. The multi-task learning model in this paper makes use of encoder parameter sharing, which enables the model’s underlying network to take into account both named entity recognition and intent classification features. The model learns the shared information between the two tasks while maintaining its unique characteristics during the decoding phase. The ALBERT pre-training language model is used to obtain word vectors containing semantic information and the bidirectional LSTM network is used for training. A comparative experiment of different models was conducted on Chinese medical questions dataset. Experimental results show that the proposed multi-task learning method outperforms the benchmark method in terms of precision, recall and F1 value. Compared with the single-task model, the generalization ability of the model has been improved.

Large-scale pre-trained language representation and its promising performance in various downstream applications have become an area of interest in the field of natural language processing (NLP). There has been huge interest in further increasing the model’s size in order to outperform the best previously obtained performances. However, at some point, increasing the model’s parameters may lead to reaching its saturation point due to the limited capacity of GPU/TPU. In addition to this, such models are mostly available in English or a shared multilingual structure. Hence, in this paper, we propose a lite BERT trained on a large corpus solely in the Romanian language, which we called “A Lite Romanian BERT (ALR-BERT)”. Based on comprehensive empirical results, ALR-BERT produces models that scale far better than the original Romanian BERT. Alongside presenting the performance on downstream tasks, we detail the analysis of the training process and its parameters. We also intend to distribute our code and model as an open source together with the downstream task.

The secularisation of the domains of the Teutonic Order in Prussia led to the establishment of the first Lutheran territorial church in the world. This fact is almost forgotten today, and this is evident even in specialised literature on the Reformation. The article outlines the introduction of the Reformation in Prussia, considering it as an example of its smooth and successful entrenchment. In order to show this, the late stage of the rule of the Teutonic Order is defined, showing that fundamental reform was triggered by a multi-layered crisis characteristic of the Order’s domains in Prussia. The article shows that, in coordination with Martin Luther and Philipp Melanchton, and assisted by his bishops, after becoming the first Duke of Prussia in 1525, Albert, the Grand Master of the Teutonic Order, implemented reforms in his domains that resembled the main problems raised by the Reformation in an almost exemplary way. But at the same time, it shows that the introduction of the Reformation in Prussia was not a unidirectional process, for Duke Albert supported Andreas Osiander’s ideas for some time, before he gradually entered the ranks of the confessors of Augsburg.

This paper introduces a series of experiments with an ALBERT over match-LSTM network on the top of pre-trained word vectors, for accurate classification of intelligent question answering and thus the guarantee of precise information service. To improve the performance of data classification, a short text classification method based on an ALBERT and match-LSTM model was proposed to overcome the limitations of the classification process, such as few vocabularies, sparse features, large amount of data, lots of noise and poor normalization. In the model, Jieba word segmentation tools and agricultural dictionary were selected to text segmentation, GloVe algorithm was then adopted to expand the text characteristic and weighted word vector according to the text of key vector, bi-directional gated recurrent unit was applied to catch the context feature information and multi-convolutional neural networks were finally established to gain local multidimensional characteristics of text. Batch normalization, Dropout, Global Average Pooling and Global Max Pooling were utilized to solve overfitting problem. The results showed that the model could classify questions accurately, with a precision of 96.8%. Compared with other classification models, such as multi-SVM model and CNN model, ALBERT+match-LSTM had obvious advantages in classification performance in intelligent Agri-tech information service.

Stock price prediction can be made more efficient by considering the price fluctuations and understanding people’s sentiments. A limited number of models understand financial jargon or have labelled datasets concerning stock price change. To overcome this challenge, we introduced FinALBERT, an ALBERT based model trained to handle financial domain text classification tasks by labelling Stocktwits text data based on stock price change. We collected Stocktwits data for over ten years for 25 different companies, including the major five FAANG (Facebook, Amazon, Apple, Netflix, Google). These datasets were labelled with three labelling techniques based on stock price changes. Our proposed model FinALBERT is fine-tuned with these labels to achieve optimal results. We experimented with the labelled dataset by training it on traditional machine learning, BERT, and FinBERT models, which helped us understand how these labels behaved with different model architectures. Our labelling method’s competitive advantage is that it can help analyse the historical data effectively, and the mathematical function can be easily customised to predict stock movement.

I defend the idea that the fact that no system is entirely isolated (“Interventionism”) can be used to explain the successful use of the microcanonical distribution in statistical mechanics. The argument turns on claims about what is needed for an adequate explanation of this fact: I argue in particular that various competing explanations do not meet reasonable conditions of adequacy, and that the most striking lacuna in Interventionism—its failure to explain the “arrow of time”—is no real defect.

The industrial fabric of north-eastern France paid a heavy tribute to the First World War. In the eastern part of the Somme, which was crossed by the front line throughout the duration of the war, many industrial sites were destroyed or badly damaged. One case is the firm of Frédéric Ehrestmann and brothers, a machine-tool manufacturer based at Albert, a city that was occupied by the Germans in 1914 and again in the spring of 1918 and which suffered heavy bombing throughout the war. A file on this factory presently held by the national archives at Pierrefitte, contains documents—official forms, written reports, plans, photos—which give a fairly complete account of the damage suffered. Three factory sites, a foundry, production equipment, a locomotive acquired in July 1914, requisitions, supplies… all are evoked following different losses. The question of the sale of equipment in 1915 and of the reconstruction of the factory in 1917 (destroyed again in 1918), throw light on the running of an industrial enterprise situated close to the front line. The explanations that the owners of the firm give concerning their decisions and choices constitute an exceptional record of an industry which has disappeared and of the professional life of industrialists in the early twentieth century.

Commentary on the life achievements of Albert Hofmann (1906–2008), one of UZH's most honored alumni.