Bag-of-words Research Articles

PhoBERT: Application in Disease Classification based on Vietnamese Symptom Analysis

Abstract Besides the successful use of support software in cutting-edge medical procedures, the significance of determining a disease early signs and symptoms before its detection is a growing pressing requirement to raise the standard of medical examination and treatment. This creates favourable conditions, reduces patient inconvenience and hospital overcrowding. Before transferring patients to an appropriate doctor, healthcare staff must have the patient’s symptoms. This study leverages the PhoBERT model to assist in classifying patients with text classification tasks based on symptoms they provided in the first stages of Vietnamese hospital admission. The outcomes of PhoBERT on more than 200 000 text-based symptoms collected from Vietnamese hospitals can improve the classification performance compared to Bag of Words (BOW) with classic machine learning algorithms, and some considered deep learning architectures such as 1D-Convolutional Neural Networks and Long Short-Term Memory. The proposed method can achieve promising results to be deployed in automatic hospital admission procedures in Vietnam.

An Adversarial Attack Analysis on Malicious Advertisement URL Detection Framework

Malicious advertisement URLs pose a security risk since they are the source of cyber-attacks, and the need to address this issue is growing in both industry and academia. Several attempts have been made in recent years for malicious URL detection using machine learning (ML). The most widely used techniques extract linguistic features of URL string to extract features like bag-of-words (BoW) before applying ML model. Existing malicious URL detection techniques require effective manual feature engineering that can handle unseen features and generalise to test data. In this study, we extract a novel set of lexical and web-scrapped features and employ ML techniques for fraudulent advertisement URL detection. The combination set of six different kinds of features precisely overcomes the obfuscation in fraudulent URL classification. Based on distinct statistical properties, we use twelve differently formatted datasets for detection, prediction and classification task. We extend our prediction analysis for mismatched and unlabelled datasets. For this framework, we analyze the performance of four ML techniques: Random Forest, Gradient Boost, XGBoost and AdaBoost in the detection part. With our proposed method, we achieve a false negative rate up to 0.0037 while maintaining high detection accuracy of 99.63%. Moreover, we employ an unsupervised learning technique for data clustering using the K-Means algorithm for the visual analysis. This paper analyses the vulnerability of decision tree-based models using the limited knowledge attack scenario. We considered the exploratory attack during the test phase and implemented Zeroth Order Optimization adversarial attack on the detection models.

Conflict and Conflicting News Discourses: An Analysis of Newspaper Coverage of Pulwama Attack

ABSTRACT Discourse analysis is used to investigate print media coverage of the Pulwama attack of 2019, a recent development in the Kashmir conflict between India and Pakistan that pushed the nuclear-armed Asian neighbours to the brink of war. To identify the news discourses, 686 headlines of news articles published in six popular newspapers on the attack and its aftermath during February 15–25, 2019, are analysed. Three levels of discourse analysis are employed: Lexical choices, social languages and conversations. Under lexical choices, the headlines are analysed to identify the bags of words used to represent places and actions, along with the literary-rhetorical devices employed. With the social languages tool, the assumed identities and intentions of the headline writer are deduced, while the conversations tool is used to identify and categorise the events covered, issues raised, actors and institutions involved and their arguments in the headlines. Five distinct discourses—News, Political, Critical, Conflict and Peace—are thus identified and characterised. Further, consulting critical discourse theories, the study findings are used to discuss the interplay of media, discourses and power, and theorise on the sociopolitical powerplays that regulate those journalistic discourses, discussing implications and the way forward for the journalists and journalism practice covering conflicts.

A pose estimation method of space non-cooperative target based on ORBFPFH SLAM

In this paper, to improve the pose measurement accuracy of the Time-of-Flight (ToF) camera for space non-cooperative targets, a pose estimation method based on ORBFPFH Simultaneous Localization and Mapping (SLAM) with effective integration of strength and depth measurement information is proposed. The primary process of the method is as follows: First, a unique ORBFPFH Bag of Words (BoW) model for space targets is trained based on the ToF image dataset of spacecraft; Secondly, the pose of the non-cooperative space target is tracked based on the ORBFPFH feature and optimized using the pose graph; Then, based on the ORBFPFH BoW model, loop closure is detected to reduce the cumulative error. Finally, the proposed method is tested on the ToF image dataset: compared with the advanced ORB-SLAM2 algorithm, the proposed method has a minor translation error and rotation error on the test data, with a mean translation error less than 0.144 m and mean rotation error less than 0.642°. The test results show that the proposed method can improve the pose estimation accuracy of space non-cooperative targets, achieve a better 3D point cloud reconstruction effect, and provide technical support for space applications such as rendezvous and docking of non-cooperative spacecraft. © 2001 Elsevier Science. All rights reserved

Texts Classification with the usage of Neural Network based on the Word2vec’s Words Representation

Assigning the submitted text to one of the predetermined categories is required when dealing with application-oriented texts. There are many different approaches to solving this problem, including using neural network algorithms. This article explores using neural networks to sort news articles based on their category. Two word vectorization algorithms are being used — The Bag of Words (BOW) and the word2vec distributive semantic model. For this work the BOW model was applied to the FNN, whereas the word2vec model was applied to CNN. We have measured the accuracy of the classification when applying these methods for ad texts datasets. The experimental results have shown that both of the models show us quite the comparable accuracy. However, the word2vec encoding used for CNN showed more relevant results, regarding to the texts semantics. Moreover, the trained CNN, based on the word2vec architecture, has produced a compact feature map on its last convolutional layer, which can then be used in the future text representation. I.e. Using CNN as a text encoder and for learning transfer.

Incremental Bag of Words with Gradient Orientation Histogram for Appearance-Based Loop Closure Detection

This paper proposes a novel approach for appearance-based loop closure detection using incremental Bag of Words (BoW) with gradient orientation histograms. The presented approach involves dividing and clustering image blocks into local region features and representing them using gradient orientation histograms. To improve the efficiency of the loop closure detection process, the vocabulary Clustering Feature (CF) tree is generated and updated in real time using the Balanced Iterative Reducing and Clustering using Hierarchies (BIRCH) algorithm, which is combined with an inverted index for the efficient selection of candidates and calculation of similarity. Moreover, temporally close and highly similar images are grouped to generate islands, which enhances the accuracy and efficiency of the loop closure detection process. The proposed approach is evaluated on publicly available datasets, and the results demonstrate high recall and precision.

The Study of Fishing Vessel Behavior Identification Based on AIS Data: A Case Study of the East China Sea

The goal of this paper is to strengthen the supervision of fishing behavior in the East China Sea and effectively ensure the sustainable development of fishery resources. Based on AIS data, this paper analyzes three types of fishing boats (purse seine operation, gill net operation and trawl operation) and uses the cubic spline interpolation algorithm to optimize the ship trajectory and construct high-dimensional features. It proposes a new coding method for fishing boat trajectory sequences. This method uses the Geohash algorithm to divide the East China Sea into grids and generate corresponding numbers. Then, the ship trajectory is mapped to the grid, the fishing boat trajectory points are associated with the divided grid, and the ship trajectory ID is extracted from the corresponding grid. The extracted complete trajectory sequence passes through the CBOW (continuous bag of words) model, and the correlation of trajectory points is fully learned. Finally, the fishing boat trajectory is converted from coordinate sequence to trajectory vector, and the processed trajectory sequence is trained by the LightGBM algorithm. In order to obtain the optimal classification effect, the optimal superparameter combination is selected. We put forward a LightGBM algorithm based on the Bayesian optimization algorithm, and obtained the classification results of three kinds of fishing boats. The final result was evaluated using the F1_score. Experimental results show that the F1_score trained with the proposed trajectory vectorization method is the highest, with a training accuracy of 0.925. Compared to XgBoost and CatBoost, the F1_score increased by 1.8% and 1.2%, respectively. The results show that this algorithm demonstrates strong applicability and effectiveness in fishery area evaluations and is significant for strengthening fishery resource management.

End-to-End page-Level assessment of handwritten text recognition

The evaluation of Handwritten Text Recognition (HTR) systems has traditionally used metrics based on the edit distance between HTR and ground truth (GT) transcripts, at both the character and word levels. This is very adequate when the experimental protocol assumes that both GT and HTR text lines are the same, which allows edit distances to be independently computed to each given line. Driven by recent advances in pattern recognition, HTR systems increasingly face the end-to-end page-level transcription of a document, where the precision of locating the different text lines and their corresponding reading order (RO) play a key role. In such a case, the standard metrics do not take into account the inconsistencies that might appear. In this paper, the problem of evaluating HTR systems at the page level is introduced in detail. We analyse the convenience of using a two-fold evaluation, where the transcription accuracy and the RO goodness are considered separately. Different alternatives are proposed, analysed and empirically compared both through partially simulated and through real, full end-to-end experiments. Results support the validity of the proposed two-fold evaluation approach. An important conclusion is that such an evaluation can be adequately achieved by just two simple and well-known metrics: the Word Error Rate (WER), that takes transcription sequentiality into account, and the here re-formulated Bag of Words Word Error Rate (bWER), that ignores order. While the latter directly and very accurately assess intrinsic word recognition errors, the difference between both metrics (ΔWER) gracefully correlates with the Normalised Spearman’s Foot Rule Distance (NSFD), a metric which explicitly measures RO errors associated with layout analysis flaws. To arrive to these conclusions, we have introduced another metric called Hungarian Word Word Rate (hWER), based on a here proposed regularised version of the Hungarian Algorithm. This metric is shown to be always almost identical to bWER and both bWER and hWER are also almost identical to WER whenever HTR transcripts and GT references are guarantee to be in the same RO.

Deep bidirectional LSTM for disease classification supporting hospital admission based on pre-diagnosis: a case study in Vietnam.

Overcrowding in hospitals in Vietnam has caused many disadvantages in receiving and treating patients. Especially at the stage of receiving and diagnosing procedures taking patients to the treatment departments in the hospital takes up much time. This study proposes a text-based disease diagnosis using text processing techniques (such as Bag of Words, Term Frequency- Inverse Document Frequency, and Tokenizer) combined with classifiers (such as Random Forests (RF), Multi-Layer Perceptron (MLP), Embeddings and Bidirectional Long Short-term memory (LSTM)) on symptoms. As observed from the results, deep Bidirectional LSTM can reach 0.982 in AUC in the classification of 10 diseases on 230,457 samples of pre-diagnosis collected from Vietnam hospitals used in the training and testing phases. The proposed approach is expected to provide a way to automate patient flow in hospitals to improve healthcare in the future.

Class Balancing Methods Comparison for Software Requirements Classification on Support Vector Machines

Cost, time, and development effort can increase due to errors in analyzing functional and non-functional software requirements. To minimize these errors, previous research has tried to classify software requirements, especially non-functional requirements, on the PROMISE dataset using the Bag of Words (BoW) feature extraction and the Support Vector Machine (SVM) classification algorithm. On the other hand, the unbalanced distribution of class labels tends to decrease the evaluation result. Moreover, most software requirements are usually functional requirements. Therefore, there is a tendency for classifier models to classify test data as functional requirements. Previous research has performed class balancing on a dataset to handle unbalanced data. The study can achieve better classification evaluation results. Based on the previous research, this study proposes to combine the class balancing method and the SVM algorithm. K-fold cross-validation is used to optimize the training and test data to be more consistent in developing the SVM model. Tests were carried out on the value of K in k-fold, i.e., 5, 10, and 15. Results are measured by accuracy, f1-score, precision, and recall. The Public Requirements (PURE) dataset has been used in this research. Results show that SVM with class balancing can classify software requirements more accurately than SVM without class balancing. Random Over Sampling is the class balancing method with the highest evaluation score for classifying software requirements on SVM. The results showed an improvement in the average value of accuracy, f1 score, precision, and recall in SVM by 22.07%, 19.67%, 17.73%, and 19.67%, respectively.

Automated Classification of Free-Text Radiology Reports: Using Different Feature Extraction Methods to Identify Fractures of the Distal Fibula.

Radiology reports mostly contain free-text, which makes it challenging to obtain structured data. Natural language processing (NLP) techniques transform free-text reports into machine-readable document vectors that are important for creating reliable, scalable methods for data analysis. The aim of this study is to classify unstructured radiograph reports according to fractures of the distal fibula and to find the best text mining method. We established a novel German language report dataset: a designated search engine was used to identify radiographs of the ankle and the reports were manually labeled according to fractures of the distal fibula. This data was used to establish a machine learning pipeline, which implemented the text representation methods bag-of-words (BOW), term frequency-inverse document frequency (TF-IDF), principal component analysis (PCA), non-negative matrix factorization (NMF), latent Dirichlet allocation (LDA), and document embedding (doc2vec). The extracted document vectors were used to train neural networks (NN), support vector machines (SVM), and logistic regression (LR) to recognize distal fibula fractures. The results were compared via cross-tabulations of the accuracy (acc) and area under the curve (AUC). In total, 3268 radiograph reports were included, of which 1076 described a fracture of the distal fibula. Comparison of the text representation methods showed that BOW achieved the best results (AUC = 0.98; acc = 0.97), followed by TF-IDF (AUC = 0.97; acc = 0.96), NMF (AUC = 0.93; acc = 0.92), PCA (AUC = 0.92; acc = 0.9), LDA (AUC = 0.91; acc = 0.89) and doc2vec (AUC = 0.9; acc = 0.88). When comparing the different classifiers, NN (AUC = 0,91) proved to be superior to SVM (AUC = 0,87) and LR (AUC = 0,85). An automated classification of unstructured reports of radiographs of the ankle can reliably detect findings of fractures of the distal fibula. A particularly suitable feature extraction method is the BOW model. · The aim was to classify unstructured radiograph reports according to distal fibula fractures.. · Our automated classification system can reliably detect fractures of the distal fibula.. · A particularly suitable feature extraction method is the BOW model.. · Dewald CL, Balandis A, Becker LS et al. Automated Classification of Free-Text Radiology Reports: Using Different Feature Extraction Methods to Identify Fractures of the Distal Fibula. Fortschr Röntgenstr 2023; 195: 713 - 719.

Hybrid classifier for sentiment analysis in Malayalam with modified TF-IDF features

Sentiment Analysis (SA) is a computational study that examines people’s opinions, attitudes, and opinions based on their written text. Keralites’ mother tongue, Malayalam, is the most often used language to express themselves on Twitter. As there is no automatic Sentiment Analyzer in Malayalam, the SA of Twitter messages is necessary. In this research work, a Malayalam sentiment analysis model is introduced. The input raw data in the form of reviews are fed for the pre-processing stage. The pre-processing module includes sentence tokenization via Sandi Splitting and parts of speech (POS) tagging. Subsequently, Bag of Word, the Proposed weightage-based Term Frequency-Inverse Document Frequency, Unigram with the dictionary and Unigram with dictionary including negation words have been considered for feature vector formation. Finally, the review classification is undergone via the proposed hybrid classifier, which is constructed by hybridizing the Long Short-Term Memory (LSTM) and Convolutional Neural Network (CNN), respectively. To enhance the classification performance, the weight of the CNN classifier will be fine-tuned via an Improved Bumble bee mating Optimization (IBBMO), which is an advanced version of standard BBMO. The performance of the proposed work is compared over other conventional models in terms of positive, negative, convergence, and other measures. The accuracy of the proposed method is 95.436, which is much better than the existing works like DBN[Formula: see text]86.25, NN[Formula: see text]87.080, SVM[Formula: see text]87.234, WOA+Hybrid Classifier[Formula: see text]88.440, LA+Hybrid Classifier[Formula: see text]89.35285, CSO+Hybrid Classifier[Formula: see text]92.021, BBMO+Hybrid Classifier[Formula: see text]93.4248, and ML techniques[Formula: see text]94.014, at the 90th training percentage.

Three-Dimensional Lidar Localization and Mapping with Loop-Closure Detection Based on Dense Depth Information

This paper presents a novel lidar SLAM system for localizing a mobile robot to build a map of the environment. To identify the unknown transform matrix, we design a new scan-matching approach, in which a point cloud segmentation algorithm is additionally integrated. Different from the traditional normal distribution transform algorithm for point cloud registration, our newly proposed one additionally incorporates a ground point remover and a point cloud segmentation method. By employing the point cloud segmentation algorithm to divide the point cloud space into different cells, the newly proposed algorithm can guarantee the continuity and convergence of the cost function. To tackle the recognition difficulties that the camera-based loop-closure detection heavily depends on the environment’s appearance, a depth-completion algorithm is introduced to fuse sensor data to ensure the robustness of the algorithm. Moreover, the bags of binary words (DBoW) are adopted to improve the image-matching quality. Finally, experimental results are presented to illustrate the effectiveness of the proposed system.

BoW3D: Bag of Words for Real-Time Loop Closing in 3D LiDAR SLAM

Loop closing is a fundamental part of simultaneous localization and mapping (SLAM) for autonomous mobile systems. In the field of visual SLAM, bag of words (BoW) has achieved great success in loop closure. The BoW features for loop searching can also be used in the subsequent 6-DoF loop correction. However, for 3D LiDAR SLAM, the state-of-the-art methods may fail to effectively recognize the loop in real time, and usually cannot correct the full 6-DoF loop pose. To address this limitation, we present a novel Bag of Words for real-time loop closing in 3D LiDAR SLAM, called BoW3D. Our method not only efficiently recognizes the revisited loop places, but also corrects the full 6-DoF loop pose in real time. BoW3D builds the bag of words based on the 3D LiDAR feature LinK3D, which is efficient, pose-invariant and can be used for accurate point-to-point matching. We furthermore embed our proposed method into 3D LiDAR odometry system to evaluate loop closing performance. We test our method on public dataset, and compare it against other state-of-the-art algorithms. BoW3D shows better performance in terms of F1 max and extended precision scores on most scenarios. It is noticeable that BoW3D takes an average of 48 ms to recognize and correct the loops on KITTI 00 (includes 4K+ 64-ray LiDAR scans), when executed on a notebook with an Intel Core i7 @2.2 GHz processor. We release the implementation of our method here: https://github.com/YungeCui/BoW3D.

Detecting AI-generated essays: the ChatGPT challenge

PurposeWith the advent of ChatGPT, a sophisticated generative artificial intelligence (AI) tool, maintaining academic integrity in all educational settings has recently become a challenge for educators. This paper discusses a method and necessary strategies to confront this challenge.Design/methodology/approachIn this study, a language model was defined to achieve high accuracy in distinguishing ChatGPT-generated essays from human written essays with a particular focus on “not falsely” classifying genuinely human-written essays as AI-generated (Negative).FindingsVia support vector machine (SVM) algorithm 100% accuracy was recorded for identifying human generated essays. The author discussed the key use of Recall and F2 score for measuring classification performance and the importance of eliminating False Negatives and making sure that no actual human generated essays are incorrectly classified as AI generated. The results of the proposed model's classification algorithms were compared to those of AI-generated text detection software developed by OpenAI, GPTZero and Copyleaks.Practical implicationsAI-generated essays submitted by students can be detected by teachers and educational designers using the proposed language model and machine learning (ML) classifier at a high accuracy. Human (student)-generated essays can and must be correctly identified with 100% accuracy even if the overall classification accuracy performance is slightly reduced.Originality/valueThis is the first and only study that used an n-gram bag-of-words (BOWs) discrepancy language model as input for a classifier to make such prediction and compared the classification results of other AI-generated text detection software in an empirical way.

Investigating the Extent to which Distributional Semantic Models Capture a Broad Range of Semantic Relations.

Distributional semantic models (DSMs) are a primary method for distilling semantic information from corpora. However, a key question remains: What types of semantic relations among words do DSMs detect? Prior work typically has addressed this question using limited human data that are restricted to semantic similarity and/or general semantic relatedness. We tested eight DSMs that are popular in current cognitive and psycholinguistic research (positive pointwise mutual information; global vectors; and three variations each of Skip-gram and continuous bag of words (CBOW) using word, context, and mean embeddings) on a theoretically motivated, rich set of semantic relations involving words from multiple syntactic classes and spanning the abstract-concrete continuum (19 sets of ratings). We found that, overall, the DSMs are best at capturing overall semantic similarity and also can capture verb-noun thematic role relations and noun-noun event-based relations that play important roles in sentence comprehension. Interestingly, Skip-gram and CBOW performed the best in terms of capturing similarity, whereas GloVe dominated the thematic role and event-based relations. We discuss the theoretical and practical implications of our results, make recommendations for users of these models, and demonstrate significant differences in model performance on event-based relations.

Novel cartographer using an OAK-D smart camera for indoor robots location and navigation

In recent years, service robots have been widely used in people’s daily life, and with the development of more and more intelligence, people put forward higher requirements for autonomous positioning and navigation functions of robots. Like outdoor navigation, indoor navigation also needs the support of navigation data. Although the indoor positioning and navigation scheme based on cameras, lidars and other sensors is gradually developing, due to the complexity of the indoor structure, manual production of indoor navigation data is time-consuming and laborious, and the efficiency is relatively low. In order to solve the problem of low productivity and improve the accuracy of robot automatic navigation, we added a new type of intelligent camera, called OpenCV AI kit or OAK-D, and proposed a method to automatically build data files that can be used for indoor navigation and location services using indoor 3D point cloud data. This intelligent camera performs neural reasoning on chips that do not use GPUs. It can also use stereo drills for depth estimation, and use 4K color camera images as input to run the neural network model. Python API can be called to realize real-time detection of indoor doors, windows and other static objects. The target detection technology uses an artificial intelligence camera, and the robot can well identify and accurately mark on the indoor map. In this paper, a high-performance indoor robot navigation system is developed, and multisensor fusion technology is designed. Environmental information is collected through artificial intelligent camera (OAK-D), laser lidar, and data fusion is carried out. In the experiment part of this paper,The static fusion map module is created based on the laser sensor information and the sensor information of the depth camera, the hierarchical dynamic cost map module is created in the real-time navigation, and the global positioning of the robot is realized by combining the word bag model and the laser point cloud matching. Then a software system is realized by integrating each module. The experiment proves that the system is practical and effective, and has practical value.

Context awareness based Sketch-DeepNet architecture for hand-drawn sketches classification and recognition in AIoT.

A sketch is a black-and-white, 2-D graphical representation of an object and contains fewer visual details as compared to a colored image. Despite fewer details, humans can recognize a sketch and its context very efficiently and consistently across languages, cultures, and age groups, but it is a difficult task for computers to recognize such low-detail sketches and get context out of them. With the tremendous increase in popularity of IoT devices such as smartphones and smart cameras, etc., it has become more critical to recognize free hand-drawn sketches in computer vision and human-computer interaction in order to build a successful artificial intelligence of things (AIoT) system that can first recognize the sketches and then understand the context of multiple drawings. Earlier models which addressed this problem are scale-invariant feature transform (SIFT) and bag-of-words (BoW). Both SIFT and BoW used hand-crafted features and scale-invariant algorithms to address this issue. But these models are complex and time-consuming due to the manual process of features setup. The deep neural networks (DNNs) performed well with object recognition on many large-scale datasets such as ImageNet and CIFAR-10. However, the DDN approach cannot be carried out for hand-drawn sketches problems. The reason is that the data source is images, and all sketches in the images are, for example, 'birds' instead of their specific category (e.g., 'sparrow'). Some deep learning approaches for sketch recognition problems exist in the literature, but the results are not promising because there is still room for improvement. This article proposed a convolutional neural network (CNN) architecture called Sketch-DeepNet for the sketch recognition task. The proposed Sketch-DeepNet architecture used the TU-Berlin dataset for classification. The experimental results show that the proposed method beats the performance of the state-of-the-art sketch classification methods. The proposed model achieved 95.05% accuracy as compared to existing models DeformNet (62.6%), Sketch-DNN (72.2%), Sketch-a-Net (77.95%), SketchNet (80.42%), Thinning-DNN (74.3%), CNN-PCA-SVM (72.5%), Hybrid-CNN (84.42%), and human recognition accuracy of 73% on the TU-Berlin dataset.

Machine learning: Classifiying taxpayer’s supervising zone based on the street address using Natural Language Processing algorithm

Assigning taxpayers into their respective Account Representatives is a crucial step to optimize Taxpayers supervision. However, the large number of registered taxpayers and missing data has been a great challenge. A lot of taxpayers only include their street addresses and no additional information such as RT, RW, etc. This will cause additional work to manually search each taxpayer address in the internet and manually assigned, which is not efficient and takes a lot of time. This study will try to solve this problem using Natural Language Processing algorithm. Efficiency and accuracy are the key on creating machine learning model. Choosing the right classifier is crucial to the accuracy. Other than the classifier, managing text data is also challenging, since it cannot be understood directly by computers. Thus, this study will also include how we could transform the text data into arrays of numbers called Bag of Words.

Lexical Ambiguity in Political Rhetoric: Why Morality Doesn't Fit in a Bag of Words

AbstractHow do politicians use moral appeals in their rhetoric? Previous research suggests that morality plays an important role in elite communication and that the endorsement of specific values varies systematically across the ideological spectrum. We argue that this view is incomplete since it only focuses on whether certain values are endorsed and not how they are contextualized by politicians. Using a novel sentence embedding approach, we show that although liberal and conservative politicians use the same moral terms, they attach diverging meanings to these values. Accordingly, the politics of morality is not about the promotion of specific moral values per se but, rather, a competition over their respective meaning. Our results highlight that simple dictionary-based methods to measure moral rhetoric may be insufficient since they fail to account for the semantic contexts in which words are used and, therefore, risk overlooking important features of political communication and party competition.