Semantic Relations Research Articles

MRME-Net: Towards multi-semantics learning and long-tail problem of efficient event detection from social messages

Discovering trending social events (e.g., major meetings, political scandals, natural disasters, etc.) from social messages is vital because it emphasizes important events and can help people comprehend the world. However, the heterogeneous semantics enrichment, severe long-tail problems, and sparse text contents of social messages pose great challenges to event detection, often leading to limited generalization ability and accuracy. In this paper, we propose a novel Multi-Relational Meta-Enhanced Network (MRME-Net) architecture to learn social events. First, we model social messages into a multi-relational message graph, incorporating abundant meta-semantics along with various meta-relations. Second, we present a multi-relational graph attention network based on Sophia by using a dual-step message aggregation mechanisms to capture the local features of neighboring messages and global semantics of mutiple relations and ultimately learn social message embeddings. We use Sophia optimizer to reduce the massive time and cost of training. Third, in order to address the long-tail problem, we introduce a locally-adapted meta-learning framework in social event detection for the first time and propose a novel META-TAILENH embedding enhancement strategy to refine tail node embeddings in multi-relational graph. Eventually, we conduct the detection of social events according to the hierarchical clustering algorithm. Extensive experiments have been carried out to evaluate MRME-Net on the MAVEN and Twitter dataset, revealing a notable improvement of 3 %–13 %, 4 %–20 % and 6 %–30 % increases on NMI, AMI and ARI in the social event detection task.

MRFTrans: Multimodal Representation Fusion Transformer for monocular 3D semantic scene completion

The complete understanding of 3D scenes is crucial in robotic visual perception, impacting tasks such as motion planning and map localization. However, due to the limited field of view and scene occlusion constraints of sensors, inferring complete scene geometry and semantic information from restricted observations is challenging. In this work, we propose a novel Multimodal Representation Fusion Transformer framework (MRFTrans) that robustly fuses semantic, geometric occupancy, and depth representations for monocular-image-based scene completion. MRFTrans centers on an affinity representation fusion transformer, integrating geometric occupancy and semantic relationships within a transformer architecture. This integration enables the modeling of long-range dependencies within scenes for inferring missing information. Additionally, we present a depth representation fusion method, efficiently extracting reliable depth knowledge from biased monocular estimates. Extensive experiments demonstrate MRFTrans’s superiority, setting a new benchmark on SemanticKITTI and NYUv2 datasets. It significantly enhances completeness and accuracy, particularly in large structures, movable objects, and scene components with major occlusions. The results underscore the benefits of the affinity-aware transformer and robust depth fusion in monocular-image-based completion.

Lexical-semantic activation in French-Spanish or French-English bilingual toddlers: An event-related potential (ERP) investigation

Both behavioral and neurophysiological evidence shows that lexical-semantic organization emerges by two years in monolingual children. Research in bilingual children is more scarce, and there is only a limited amount of neurophysiological evidence of the effect of language dominance on lexical-semantic activation. In the present event-related potential (ERP) study, we investigated whether bilingual French-Spanish and French-English learning 24- to 30-month-olds activate semantic relations between words similarly in their both languages, and whether the priming effects are similar in children learning two different language pairs. Participants were presented with related and unrelated dominant and non-dominant language word pairs in a within-language lexical-semantic priming paradigm. The amplitudes of N400 were modulated by trial type, language dominance and language group. A language-independent priming effect - more pronounced N400 amplitudes for unrelated than for related target words - was found in the group of toddlers learning French and Spanish. In the group of toddlers learning French and English, a priming effect was observed only in their non-dominant language. Our results propose that the language pair may contribute to lexical-semantic facilitation in priming tasks during early childhood.

Evolving malware detection through instant dynamic graph inverse reinforcement learning

The rapid development and increasing evolution of malware necessitate novel defensive techniques with high accuracy and minimal false positives to safeguard information systems against potential threats. Unfortunately, current malware detection methods, primarily relying on deep learning to identify malicious fingerprints from extensive training sets, prove ineffective against few-shot and evolving malware variants. To address these challenges, this paper introduces MalIRL, which designs a model-free inverse reinforcement learning (IRL) mechanism to automatically capture the evolving attack intent of malware. Specifically, MalIRL explores six representative categories of malware actions and employs sliding windows to organically divide the massive malware execution event stream into multiple attack stages, achieving a reduced action space and state space. To model dynamic malicious environments, MalIRL proposes an instant dynamic heterogeneous graph representation learning technique. This technique learns state representations of different malware attack stages, enhancing detection accuracy and efficiency by incrementally capturing the newly added contextual semantics of diverse malware entities and relations. In experiments with three real-world malware datasets, MalIRL surpasses existing state-of-the-art methods, particularly in few-shot malware detection scenarios, MalIRL exhibits performance benefits of up to 18.9%.

A Study of Appraisal System and Its Texture/Textuality Mechanisms

Language, as a social symbol, functions to express world experiences, construct social role relationships, and choose words and sentences to organize discourse. The interaction between language meta-functions is a problem deserved studying. From the perspective of textual meta-function, this paper tries to discuss the mechanisms of organizing Appraisal system as a texture, and construct a rather comprehensive analysis framework of the textual devices for evaluation meaning as a whole text. The framework built consists of three strata (contextual, discourse-semantic, and lexico-grammatical), two dimensions (micro and macro), and an intermediary variable (rhetorical strategies). At the contextual stratum, the generic structure of genre and mode of register, through rhetorical strategies, determine the distribution and combination patterns of evaluative meaning at the discourse-semantic stratum. In micro dimension, cohesion, information distribution pattern, and evaluative prosody, together with theme, new information and logical semantic relations in macro dimension, are combined to organize evaluation meaning at the discourse-semantic stratum. At the lexico-grammatical stratum, Appraisal meaning is thus embodied and realized by the evaluative lexical chains, thematic structures, new information structures, connectives and meta-discourse markers at the discourse-semantic level of a text. This framework reflects the interactions between the interpersonal meta-function and textual meta-function of Appraisal system and provides insights for the development of Appraisal Theory.

The role of culture and semantic organization in working memory updating

ABSTRACT Westerners tend to relate items in a categorical manner, whereas Easterners focus more on functional relationships. The present study extended research on semantic organization in long-term memory to working memory. First, Americans’ and Turks’ preferences for categorical versus functional relationships were tested. Second, working memory interference was assessed using a 2-back working memory paradigm in which lure items were categorically and functionally related to targets. Next, a mediation model tested direct effects of culture and semantic organization on working memory task behaviour, and the indirect effect, whether semantic organization mediated the relationship between culture and working memory interference. Whereas Americans had slower response times to correctly rejecting functional lures compared to categorical lures, conditions did not differ for Turks. However, semantic organization did not mediate cultural difference in working memory interference. Across cultures, there was evidence that semantic organization affected working memory errors, with individuals who endorsed categorical more than functional pairings committing more categorical than functional errors on the 2-back task. Results align with prior research suggesting individual differences in use of different types of semantic relationships, and further that literature by indicating effects on interference in working memory. However, these individual differences may not be culture-dependent.

Research on Aspect-Level Sentiment Analysis Based on Adversarial Training and Dependency Parsing

Aspect-level sentiment analysis is used to predict the sentiment polarity of a specific aspect in a sentence. However, most current research cannot fully utilize semantic information, and the models lack robustness. Therefore, this article proposes a model for aspect-level sentiment analysis based on a combination of adversarial training and dependency syntax analysis. First, BERT is used to transform word vectors and construct adjacency matrices with dependency syntactic relationships to better extract semantic dependency relationships and features between sentence components. A multi-head attention mechanism is used to fuse the features of the two parts, simultaneously perform adversarial training on the BERT embedding layer to enhance model robustness, and, finally, to predict emotional polarity. The model was tested on the SemEval 2014 Task 4 dataset. The experimental results showed that, compared with the baseline model, the model achieved significant performance improvement after incorporating adversarial training and dependency syntax relationships.

Temporal Scene Montage for Self-Supervised Video Scene Boundary Detection

Once a video sequence is organized as basic shot units, it is of great interest to temporally link shots into semantic-compact scene segments to facilitate long video understanding. However, it still challenges existing video scene boundary detection methods to handle various visual semantics and complex shot relations in video scenes. We proposed a novel self-supervised learning method, Video Scene Montage for Boundary Detection (VSMBD), to extract rich shot semantics and learn shot relations using unlabeled videos. More specifically, we present Video Scene Montage (VSM) to synthesize reliable pseudo scene boundaries, which learns task-related semantic relations between shots in a self-supervised manner. To lay a solid foundation for modeling semantic relations between shots, we decouple visual semantics of shots into foreground and background. Instead of costly learning from scratch as in most previous self-supervised learning methods, we build our model upon large-scale pre-trained visual encoders to extract the foreground and background features. Experimental results demonstrate VSMBD trains a model with strong capability in capturing shot relations, surpassing previous methods by significant margins. The code is available at https://github.com/mini-mind/VSMBD.

SEMANTICS OF THE CONCEPT OF “STEED”/“HORSE”IN THE RUSSIAN ART OF THE FIRST THIRD OF THE20TH CENTURY: “A HORSE AS A HORSE” AND “WE AREALL A LITTLE BIT HORSES”

The article continues the research on the artistic image in the Russian art of the first third of the 20th century and aims at studying the concepts of “steed”/“horse” in the art of the 1900s-1920s. From the beginning of revolutionary transformations in Russian art to the changes after 1917, the semantic meaning of the artistic image in its evolution is of particular interest. The basis for the study is largely fiction, in which the horse image was formed most fully. The multidimensionality of semantic relations within the boundaries of the synonymic paradigm with the general meaning of “horse” is examined in the article on the material of fine arts. The author studies the diversity of artistic roles, symbolic and semantic functions, which were given to the horse image in the art of a certain period and reveals the specificity of these roles depending on the artistic concept. Identification and recognition of the image occurs at different levels. The content of the concept is not revealed only through the plot. Unexpected semantic associations, suggested to the artist by visual, worldly, intellectual, emotional experience or creative intuition, are also possible. In this context, the quotations in the title of the article mark two main notions: “a horse as a horse” (V. Shershenevich) actualizes the direct meaning of “a horse as an animal”, and “we are all a little bit horses” (V. Mayakovsky) is a zoometaphor defining the inner state of a person. At this stage of the study, we can distinguish five main sections that allow us to structure and order the set of works containing the horse image: a village, a city, a rider, Apocalypse, a historical event (here the Civil War and collectivization). In order to fully understand the specifics of the semantic content of the horse image in the art during the study period, it is necessary to carry out some additional research. So far, the proposed text cannot claim to be an exhaustive solution to the problem, but outlines the vectors that allow us to continue its study.

A Unified Framework for Jointly Compressing Visual and Semantic Data

The rapid advancement of multimedia and imaging technologies has resulted in increasingly diverse visual and semantic data. A large range of applications such as remote-assisted driving requires the amalgamated storage and transmission of various visual and semantic data. However, existing works suffer from the limitation of insufficiently exploiting the redundancy between different types of data. In this article, we propose a unified framework to jointly compress a diverse spectrum of visual and semantic data, including images, point clouds, segmentation maps, object attributes, and relations. We develop a unifying process that embeds the representations of these data into a joint embedding graph according to their categories, which enables flexible handling of joint compression tasks for various visual and semantic data. To fully leverage the redundancy between different data types, we further introduce an embedding-based adaptive joint encoding process and a Semantic Adaptation Module to efficiently encode diverse data based on the learned embeddings in the joint embedding graph. Experiments on the Cityscapes, MSCOCO, and KITTI datasets demonstrate the superiority of our framework, highlighting promising steps toward scalable multimedia processing.

Towards a semi-automatic classifier of malware through tweets for early warning threat detection

This paper presents a method for developing a malware ontology structure by detecting malware instances on Twitter. The ontology represents a semi-automatic classifier fed by the data extracted from tweets. In particular, the automatic part of the presented methodology relies on a pattern-based approach to detect trigger expressions leading to new information about malware, whilst the manual one covers the evaluation of the results by domain-experts, who also validate the reliability of the semantic relationships within the ontology framework. We present preliminary results on the application of our methodology to tweets extracted from MalwareBazaar database showing how the documents’ collection analysis, through Natural Language Processing (NLP) tasks, can support the knowledge retrieval and documents’ classification procedures for building early warning system of detected malware. Results obtained from this research paper within the time framework of 2023 are referred to the previous version of the current social network X.

Integrating regular expressions into neural networks for relation extraction

Relation extraction aims to identify semantic relationships between entities from the given sentences. The development of deep learning prompts a variety of neural-based relation extraction models. These models can learn features from sentences automatically, but controlling their learned patterns is equally difficult. Therefore, it is hard to take steps to improve their performance further. A possible way to modify the models is by utilizing human-defined rules to introduce rule-based knowledge and guiding the models to capture the desired features based on human intent. As an efficient human knowledge, regular expression (RE) is helpful in extracting particular targets from text, which is also logical and interpretable. A novel embedding method is proposed to transform REs into learnable representations available to neural networks (NNs). Then, the rule-based representations are investigated to be integrated into different levels of NNs. Three representative relation extraction models based on recurrent neural network (RNN), convolutional neural network (CNN), and Transformer encoder, respectively, are selected as baselines to incorporate our rule-based representations. The experimental results show that the models with rule-based information outperform their original ones on the F1-score by 3.75%, 4.06%, and 2.57%, and the identification accuracy of complex samples is also significantly improved, verifying the effectiveness of our approaches. Our work is easy to reproduce and can be generally applied to various neural networks, proving that data-driven relation extraction methods can take advantage of the rule-based knowledge provided by REs.

Cross-linguistically consistent semantic and syntactic annotation of child-directed speech

AbstractCorpora of child speech and child-directed speech (CDS) have enabled major contributions to the study of child language acquisition, yet semantic annotation for such corpora is still scarce and lacks a uniform standard. Semantic annotation of CDS is particularly important for understanding the nature of the input children receive and developing computational models of child language acquisition. For example, under the assumption that children are able to infer meaning representations for (at least some of) the utterances they hear, the acquisition task is to learn a grammar that can map novel adult utterances onto their corresponding meaning representations, in the face of noise and distraction by other contextually possible meanings. To study this problem and to develop computational models of it, we need corpora that provide both adult utterances and their meaning representations, ideally using annotation that is consistent across a range of languages in order to facilitate cross-linguistic comparative studies. This paper proposes a methodology for constructing such corpora of CDS paired with sentential logical forms, and uses this method to create two such corpora, in English and Hebrew. The approach enforces a cross-linguistically consistent representation, building on recent advances in dependency representation and semantic parsing. Specifically, the approach involves two steps. First, we annotate the corpora using the Universal Dependencies (UD) scheme for syntactic annotation, which has been developed to apply consistently to a wide variety of domains and typologically diverse languages. Next, we further annotate these data by applying an automatic method for transducing sentential logical forms (LFs) from UD structures. The UD and LF representations have complementary strengths: UD structures are language-neutral and support consistent and reliable annotation by multiple annotators, whereas LFs are neutral as to their syntactic derivation and transparently encode semantic relations. Using this approach, we provide syntactic and semantic annotation for two corpora from CHILDES: Brown’s Adam corpus (English; we annotate $$\approx$$ ≈ 80% of its child-directed utterances), all child-directed utterances from Berman’s Hagar corpus (Hebrew). We verify the quality of the UD annotation using an inter-annotator agreement study, and manually evaluate the transduced meaning representations. We then demonstrate the utility of the compiled corpora through (1) a longitudinal corpus study of the prevalence of different syntactic and semantic phenomena in the CDS, and (2) applying an existing computational model of language acquisition to the two corpora and briefly comparing the results across languages.

A case study of comparative metaphor analysis in Finnish and Hungarian news texts

Abstract Despite remarkable previous results of cognitive linguistic analysis of metaphorization in Finnish and Hungarian even in a cross-linguistic context (Huumo 2019; Máthé 2022), a systematic comparative study of the patterns of linguistic metaphors in these two languages is still to be carried out. The aim of this paper is to fill this research gap by proposing a MIPVU-inspired protocol for metaphor identification in languages with rich morphology and presenting the preliminary results of a small-scale corpus analysis. The study is organized around the following questions: (i) How can we adapt and extend the original MIPVU method for agglutinative Uralic languages like Hungarian and Finnish? (ii) How can we implement the adapted method in a collaborative parallel annotation process of sampled news texts? (iii) What are the differences in metaphorical patterns between Finnish and Hungarian? The paper details the adapted protocol of MIPVU for the Hungarian language (Simon et al. 2023), and how it can be applied to Finnish, demonstrating its applicability in a research corpus. Our findings demonstrate a relatively similar metaphorization in Hungarian and Finnish with slight differences in terms of the frequency of metaphorical expressions, their semantic relations, and the complexity of argument structure constructions.

Lexical-semantic processing in preschoolers with Developmental Language Disorder: an eye tracking study.

This study examined lexical-semantic processing in children with Developmental Language Disorder (DLD) during visually situated comprehension of real-time spoken words. Existing evidence suggests that children with DLD may experience challenges in lexical access and retrieval, as well as greater lexical competition compared to their peers with Typical Development (TD). However, the specific nature of these difficulties remains unclear. Using eye-tracking methodology, the study investigated the real-time comprehension of semantic relationships in children with DLD and their age-matched peers. The results revealed that, for relatively frequent nouns, both groups demonstrated similar comprehension of semantic relationships. Both groups favored the semantic competitor when it appeared with an unrelated visual referent. In turn, when the semantic competitor appeared with the visual referent of the spoken word, both groups disregarded the competitor. This finding shows that, although children with DLD usually present a relatively impoverished vocabulary, frequent nouns may not pose greater difficulties for them. While the temporal course of preference for the competitor or the referent was similar between the two groups, numerical, though non-significant, differences in the extension of the clusters were observed. In summary, this research demonstrates that monolingual preschoolers with DLD exhibit similar lexical access to frequent words compared to their peers with TD. Future studies should investigate the performance of children with DLD on less frequent words to provide a comprehensive understanding of their lexical-semantic abilities.

Computational Legal Studies Comes of Age

Computational analysis techniques are transforming empirical legal scholarship. Two paradigms have emerged: law-as-code, which seeks to represent legal rules in a logical, executable format; and law-as-data, which leverages quantitative analysis of legal texts to reveal patterns and insights. This article surveys these approaches, emphasizing recent developments in large language models and generative artificial intelligence (AI). Law-as-code systems have enabled applications from tax preparation software to smart contracts, but realizing the vision of fully computational law has proven challenging. Law-as-data techniques like natural language processing and machine learning have charted the semantic relationship between courts and illuminated changes in judicial culture. Generative models showcase AI's explosive progress, with impressive feats like passing the U.S. bar example, but they also highlight limitations like factual inaccuracy and interpretability issues. Hybrid approaches integrating computational law, data science, and AI offer a promising research direction. As these tools spread, legal scholars can analyze more legal data than ever before, but they must remain cognizant of challenges like biased or low-quality data and linguistic/cultural limitations. Used judiciously alongside traditional methods, computational analysis has the potential to revolutionize empirical legal studies.

An emergent ontology for digitally modeling existing buildings: examining Kyoto’s Nishiki market

PurposeThis project examines digital modeling strategies for existing buildings. In this context, it aims to question assumptions about the need for geometric accuracy and the efficacy of predefined ontologies. As a counterpoint to prevailing digital modeling strategies, this project proposes a digital modeling approach using a project-specific, emergent ontology.Design/methodology/approachNishiki Market, in Kyoto, Japan, is studied as a test case. The emergent-ontology modeling process is introduced with an initial minimal set of operations including basic fold and trim operations applicable to surfaces. As the model develops iteratively, new situations are encountered for which existing rules are insufficient. In response, the model maker’s subjective judgment is invoked to introduce new operations, and ontological rules are allowed to expand.FindingsThe emergent-ontology approach, when executed on the Nishiki Market test case, enables representation of specific architectural qualities, highlighting semantic distinctions between digitally modeled elements of real-world features. The modeling approach generated project-specific knowledge, informing disciplinary understanding. Ontological emergence enabled semantic relationships to be disclosed and newly constructed.Originality/valueThe project proposes a novel methodology using an emergent ontology for digitally modeling existing buildings. Instead of remaining within the limitations a predefined ontology, the model maker’s subjective decisions shape the model’s ongoing development. This interpretive approach allows project-specific knowledge generation while challenging prevailing assumptions about accuracy and consistency in digital models of existing buildings.

Hybrid topic modeling method based on dirichlet multinomial mixture and fuzzy match algorithm for short text clustering

Topic modeling methods proved to be effective for inferring latent topics from short texts. Dealing with short texts is challenging yet helpful for many real-world applications, due to the sparse terms in the text and the high dimensionality representation. Most of the topic modeling methods require the number of topics to be defined earlier. Similarly, methods based on Dirichlet Multinomial Mixture (DMM) involve the maximum possible number of topics before execution which is hard to determine due to topic uncertainty, and many noises exist in the dataset. Hence, a new approach called the Topic Clustering algorithm based on Levenshtein Distance (TCLD) is introduced in this paper, TCLD combines DMM models and the Fuzzy matching algorithm to address two key challenges in topic modeling: (a) The outlier problem in topic modeling methods. (b) The problem of determining the optimal number of topics. TCLD uses the initial clustered topics generated by DMM models and then evaluates the semantic relationships between documents using Levenshtein Distance. Subsequently, it determines whether to keep the document in the same cluster, relocate it to another cluster, or mark it as an outlier. The results demonstrate the efficiency of the proposed approach across six English benchmark datasets, in comparison to seven topic modeling approaches, with 83% improvement in purity and 67% enhancement in Normalized Mutual Information (NMI) across all datasets. The proposed method was also applied to a collected Arabic tweet and the results showed that only 12% of the Arabic short texts were incorrectly clustered, according to human inspection.

Between syntax and morphology: German noun+verb units

We show that graphemic variation—at least in some writing systems—can be analysed in terms of grammatical variation given a usage- based probabilistic view of the grammar-graphemics interface. Concretely, we examine a type of noun+verb unit in German, which can be written as one word or two. We argue that the variation in writing is rooted in the units’ ambiguous status in between morphology (one word) and syntax (two words). The major influencing factors are shown to be the semantic relation between the noun and the verb (argument or oblique relation) and the morphosyntactic context. In prototypically nominal contexts, a reinterpretation of the unit as a noun+noun compound is facilitated, which favours spelling as one word, while in prototypically verbal contexts, a syntactic realisation and consequently spelling as two words is preferred. We report the results of two large-scale corpus studies and a controlled production experiment to corroborate our analysis.

Listen as you wish: Fusion of audio and text for cross-modal event detection in smart cities

In the era of smart cities, the advent of the Internet of Things technology has catalyzed the proliferation of multimodal sensor data, presenting new challenges in cross-modal event detection, particularly in audio event detection via textual queries. This paper focuses on the novel task of text-to-audio grounding (TAG), aiming to precisely localize sound segments that correspond to events described in textual queries within an untrimmed audio. This challenging new task requires multi-modal (acoustic and linguistic) information fusion as well as the reasoning for the cross-modal semantic matching between the given audio and textual query. Unlike conventional methods that often overlook the nuanced interactions between and within modalities, we introduce the Cross-modal Graph Interaction (CGI) model. This innovative approach leverages a language graph to model complex semantic relationships between query words, enhancing the understanding of textual queries. Additionally, a cross-modal attention mechanism generates snippet-specific query representations, facilitating fine-grained semantic matching between audio segments and textual descriptions. A cross-gating module further refines this process by emphasizing relevant features across modalities and suppressing irrelevant information, optimizing multimodal information fusion. Our comprehensive evaluation on the Audiogrounding benchmark dataset not only demonstrates the CGI model’s superior performance over existing methods, but also underscores the significance of sophisticated multimodal interaction in improving the efficacy of TAG in smart cities.