Sarcasm in social media, frequently conveyed through the interplay of text and images, presents significant challenges for sentiment analysis and intention mining. Existing multi-modal sarcasm detection approaches have been shown to excessively depend on superficial cues within the textual modality, exhibiting limited capability to accurately discern sarcasm through subtle text–image interactions. To address this limitation, a novel framework, InterCLIP-MEP, is proposed. This framework integrates Interactive CLIP (InterCLIP), which employs an efficient training strategy to derive enriched cross-modal representations by embedding inter-modal information directly into each encoder, while using approximately 20.6 \(\times\) fewer trainable parameters compared with existing state-of-the-art (SOTA) methods. Furthermore, a Memory-Enhanced Predictor (MEP) is introduced, featuring a dynamic dual-channel memory mechanism that captures and retains valuable knowledge from test samples during inference, serving as a nonparametric classifier to enhance sarcasm detection robustness. Extensive experiments on MMSD, MMSD2.0, and DocMSU show that InterCLIP-MEP achieves SOTA performance, specifically improving accuracy by 1.08% and F1-score by 1.51% on MMSD2.0. Under distributional shift evaluation, it attains 73.96% accuracy, exceeding its memory-free variant by nearly 10% and the previous SOTA by over 15%, demonstrating superior stability and adaptability. The implementation of InterCLIP-MEP is publicly available at https://github.com/CoderChen01/InterCLIP-MEP .

Multimodal sarcasm detection identifies ironic intent by jointly analyzing text and images. It has attracted increasing attention due to its importance in understanding user-generated content on social media. However, multimodal observations are often incomplete due to data loss during transmission or collection, leading to unreliable predictions in real-world multimodal sarcasm detection scenarios. To address incomplete observations, we further propose a Reliability-Aware Dynamic Fusion (RADF) module, which predicts the reliability of the textual, visual, and interactive views from their representations, converts these reliability scores into dynamic fusion weights via a temperature-scaled softmax to control the sharpness of the weight distribution, and refines the fused features through feature-wise scaling. In this way, degraded views are suppressed while more informative views are emphasized under incomplete observations. Extensive experiments on public datasets validate the effectiveness of our approach, which consistently outperforms existing baselines.

Chinese abbreviation prediction is essential for various natural language processing tasks, including query comprehension, entity linking, and information retrieval.Existing approaches rely on sequence tagging for abbreviation prediction.However, these approaches fail to guarantee the predicted abbreviations preserve their respective meanings and only consider context information related to the entity itself.This paper proposes a novel Chinese abbreviation prediction approach using multi-feature fusion and global context.The approach initially generates multiple candidate abbreviations using a Chinese pretrained unbalanced Transformer generative model.It then selects high-quality candidates through a multi-feature optimization stage, and finally evaluates their quality within the global contextual information.Experimental results indicate that our approach surpasses other comparable approaches in abbreviation prediction.

Event detection is a fundamental task in information extraction, aiming to identify trigger words from given texts and categorize them into distinct event types.Existing approaches for event detection predominantly rely on annotating trigger words to classify events, which can lead to semantic recognition errors due to the oversimplified semantics of these trigger words.To tackle this challenge, we propose a trigger word expansionbased approach for robust contextual event detection.Specifically, we propose a framework comprising a trigger word extractor within a model for trigger word expansion classifier.This enables event detection to be conducted through trigger word expansion, enriching the semantics of trigger words.Additionally, we leverage the GPT model to generate contexts for the expanded trigger words, thereby enhancing the contextual understanding of trigger words.Extensive experiments conducted on the standard MAVEN benchmark dataset showcase the superior performance of our approach compared to state-of-the-art methods.This confirms the effectiveness of our proposed approach over existing approaches that rely on trigger word annotation.

Classification methods for streaming data are not new, but very few current frameworks address all three of the most common problems with these tasks: concept drift, noise, and the exorbitant costs associated with labeling the unlabeled instances in data streams. Motivated by this gap in the field, we developed an active learning framework based on a dual-query strategy and Ebbinghaus's law of human memory cognition. Called CogDQS, the query strategy samples only the most representative instances for manual annotation based on local density and uncertainty, thus significantly reducing the cost of labeling. The policy for discerning drift from noise and replacing outdated instances with new concepts is based on the three criteria of the Ebbinghaus forgetting curve: recall, the fading period, and the memory strength. Simulations comparing CogDQS with baselines on six different data streams containing gradual drift or abrupt drift with and without noise show that our approach produces accurate, stable models with good generalization ability at minimal labeling, storage, and computation costs.

BACKGROUND: Multidimensional protein identification technology (MudPIT)-based shot-gun proteomics has been proven to be an effective platform for functional proteomics. In particular, the various sample preparation methods and bioinformatics tools can be integrated to improve the proteomics platform for applications like target organelle proteomics. We have recently integrated a rapid sample preparation method and bioinformatics classification system for comparative analysis of plant responses to two plant hormones, zeatin and brassinosteroid (BR). These hormones belong to two distinct classes of plant growth regulators, yet both can promote cell elongation and growth. An understanding of the differences and the cross-talk between the two types of hormone responses will allow us to better understand the molecular mechanisms and to identify new candidate genes for plant engineering. RESULTS: As compared to traditional organelle proteomics, the organelle-enrichment method both simplifies the sample preparation and increases the number of proteins identified in the targeted organelle as well as the entire sample. Both zeatin and BR induce dramatic changes in signaling and metabolism. Their shared-regulated protein components indicate that both hormones may down-regulate some key components in auxin responses. However, they have shown distinct induction and suppression of metabolic pathways in mitochondria and chloroplast. For zeatin, the metabolic pathways in sucrose and starch biosynthesis and utilization were significantly changed, yet the lipid biosynthesis remained unchanged. For BR, lipid biosynthesis and β-oxidation were both down-regulated, yet the changes in sucrose and starch metabolism were minor. CONCLUSIONS: We present a rapid sample preparation method and bioinformatics classification for effective proteomics analysis of plant hormone responses. The study highlighted the largely differing response to zeatin and brassinosteroid by the metabolic pathways in chloroplast and mitochondria.

BACKGROUND: HDX mass spectrometry is a powerful platform to probe protein structure dynamics during ligand binding, protein folding, enzyme catalysis, and such. HDX mass spectrometry analysis derives the protein structure dynamics based on the mass increase of a protein of which the backbone protons exchanged with solvent deuterium. Coupled with enzyme digestion and MS/MS analysis, HDX mass spectrometry can be used to study the regional dynamics of protein based on the m/z value or percentage of deuterium incorporation for the digested peptides in the HDX experiments. Various software packages have been developed to analyze HDX mass spectrometry data. Despite the progresses, proper and explicit statistical treatment is still lacking in most of the current HDX mass spectrometry software. In order to address this issue, we have developed the HDXanalyzer for the statistical analysis of HDX mass spectrometry data using R, Python, and RPY2. IMPLEMENTATION AND RESULTS: HDXanalyzer package contains three major modules, the data processing module, the statistical analysis module, and the user interface. RPY2 is employed to enable the connection of these three components, where the data processing module is implemented using Python and the statistical analysis module is implemented with R. RPY2 creates a low-level interface for R and allows the effective integration of statistical module for data processing. The data processing module generates the centroid for the peptides in form of m/z value, and the differences of centroids between the peptides derived from apo and ligand-bound protein allow us to evaluate whether the regions have significant changes in structure dynamics or not. Another option of the software is to calculate the deuterium incorporation rate for the comparison. The two types of statistical analyses are Paired Student's t-test and the linear combination of the intercept for multiple regression and ANCOVA model. The user interface is implemented with wxpython to facilitate the data visualization in graphs and the statistical analysis output presentation. In order to evaluate the software, a previously published xylanase HDX mass spectrometry analysis dataset is processed and presented. The results from the different statistical analysis methods are compared and shown to be similar. The statistical analysis results are overlaid with the three dimensional structure of the protein to highlight the regional structure dynamics changes in the xylanase enzyme. CONCLUSION: Statistical analysis provides crucial evaluation of whether a protein region is significantly protected or unprotected during the HDX mass spectrometry studies. Although there are several other available software programs to process HDX experimental data, HDXanalyzer is the first software program to offer multiple statistical methods to evaluate the changes in protein structure dynamics based on HDX mass spectrometry analysis. Moreover, the statistical analysis can be carried out for both m/z value and deuterium incorporation rate. In addition, the software package can be used for the data generated from a wide range of mass spectrometry instruments.

BACKGROUND: Enzyme dynamics has recently been shown to be crucial for structure-function relationship. Among various structure dynamics analysis platforms, HDX (hydrogen deuterium exchange) mass spectrometry stands out as an efficient and high-throughput way to analyze protein dynamics upon ligand binding. Despite the potential, limited research has employed the HDX mass spec platform to probe regional structure dynamics of enzymes. In particular, the technique has never been used for analyzing cell wall degrading enzymes. We hereby used xylanase as a model to explore the potential of HDX mass spectrometry for studying cell wall degrading enzymes. RESULTS: HDX mass spectrometry revealed significant intrinsic dynamics for the xylanase enzyme. Different regions of the enzymes are differentially stabilized in the apo enzyme. The comparison of substrate-binding enzymes revealed that xylohexaose can significantly stabilize the enzyme. Several regions including those near the reaction centres were significantly stabilized during the xylohexaose binding. As compared to xylohexaose, xylan induced relatively less protection in the enzyme, which may be due to the insolubility of the substrate. The structure relevance of the enzyme dynamics was discussed with reference to the three dimensional structure of the enzyme. HDX mass spectrometry revealed strong dynamics-function relevance and such relevance can be explored for the future enzyme improvement. CONCLUSION: Ligand-binding can lead to the significant stabilization at both regional and global level for enzymes like xylanase. HDX mass spectrometry is a powerful high-throughput platform to identify the key regions protected during the ligand binding and to explore the molecular mechanisms of the enzyme function. The HDX mass spectrometry analysis of cell wall degrading enzymes has provided a novel platform to guide the rational design of enzymes.

Traversing Network Address Translation (NAT) for Peer-to-Peer (P2P) communication has become a hot topic recently. Compared to UDP, establishing TCP connections for hosts behind different NATs is more complex. Thus, many TCP-based applications do not address TCP traversal through NATs. Some solutions suggest using delegates to relay all communications, or tunneling TCP over UDP. However, they require a big reform to network architecture, or using a non-standard TCP/IP stack. In this paper, we present a novel idea called TCPBridge. TCPBridge converts TCP traversal to UDP traversal without modifying any binaries of the TCP-based applications. Our design can be integrated with those P2P applications which have not solved TCP traversal problem, and extends them to support direct communications between NAT hosts. It deals with the problem of TCP traversal, so as to improve the usability of applications. We have implemented TCPBridge in several existing P2P systems. Statistics prove that TCPBridge is scalable and robust, and we believe it will benefit many other existing P2P applications.