Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks266 from public data
- Directional Skip-Gram: Explicitly Distinguishing Left and Right Context for Word Embeddings380
Direction Skip-gram (DSG), a simple but effective enhancement of the skip-gram model by explicitly distinguishing left and right context in word prediction, and which outperforms others on different datasets in semantic and syntactic evaluations.
- Challenging the long tail recommendation376
Empirical experiments show that the proposed algorithms are effective to recommend long tail items and outperform state-of-the-art recommendation techniques.
- Community assessment to advance computational prediction of cancer drug combinations in a pharmacogenomic screen362
A large drug combination screen across cancer cell lines is provided to benchmark crowdsourced methods and to computationally predict drug synergies, and genomic rationale for synergy predictions are identified.
- Mapping the Research Trends by Co-word Analysis Based on Keywords from Funded Project302
A co-word method based on keywords from funded project is proposed to map the research trends and shows that Game Theory, Supply Chain Management, Complex Network, Data Mining, Optimize, Risk Management, and Data Envelopment Analysis are hot topics.
- An extensive study on pre-trained models for program understanding and generation189
The first study for natural language-programming language pre-trained model robustness via adversarial attacks is performed and it is found that a simple random attack approach can easily fool the state-of-the-art pre- trained models and thus incur security issues.
- A Hybrid Approach to Automatic Corpus Generation for Chinese Spelling Check162
This paper proposes a novel approach of constructing CSC corpus with automatically generated spelling errors, which are either visually or phonologically resembled characters, corresponding to the OCR- and ASR-based methods, respectively.
- 154
- Network-based prediction of drug–target interactions using an arbitrary-order proximity embedded deep forest143
A network-based computational framework, termed AOPEDF, an Arbitrary-Order Proximity Embedded Deep Forest approach, for prediction of DTIs achieves high accuracy in identifying molecular targets among known drugs on two external validation sets, outperforming several state-of-the-art methods.
- Topic Memory Networks for Short Text Classification140
Experimental results show that the proposed topic memory networks for short text classification with a novel topic memory mechanism to encode latent topic representations indicative of class labels outperforms state-of-the-art models on shortText classification and generates coherent topics.
- Finding hierarchy in directed online social networks130
This paper defines a measure of hierarchy in a directed online social network, and presents an efficient algorithm to compute this measure, which is validated using ground truth including Wikipedia notability score and shows how hierarchy emerges as the size of the network increases.
- Step-Audio: Unified Understanding and Generation in Intelligent Speech Interaction129
Step-Audio is introduced, the first production-ready open-source solution that achieves state-of-the-art performance in human evaluations, especially in terms of instruction following and on open-source benchmarks like LLaMA Question.
- HME112
A novel hyperbolic metric embedding (HME) model is proposed, which projects the check-in data into aHyperbolic space, which can effectively capture the underlying hierarchical structures, which are implied by the power-law distributions of user movements.
- 108
- ACRec90
This paper proposes a random walk model using three academic metrics as basics for recommending new collaborations through co-authorship in an academic network and shows that it can improve the precision, recall rate and coverage rate of recommendation, compared with other state-of-the-art approaches.
- A bibliometric review of a decade of research: Big data in business research – Setting a research agenda86
This review study depicts the development of big data in business research using bibliometric methods, such as publication counts and trends analysis, co-citation analysis, Co-authorship analysis and keywords co-occurrence analysis, to provide scholars with a systematic understanding of the development and panoramic roadmap ofbig data research in business.
- Topic-Aware Neural Keyphrase Generation for Social Media Language85
This work proposes a sequence-to-sequence (seq2seq) based neural keyphrase generation framework, enabling absent keyphrases to be created, and shows that the model significantly outperforms both extraction and generation models without exploiting latent topics.
- M 6 Doc: A Large-Scale Multi-Format, Multi-Type, Multi-Layout, Multi-Language, Multi-Annotation Category Dataset for Modern Document Layout Analysis72
A transformer-based document layout analysis method called TransDLANet is proposed, which leverages an adaptive element matching mechanism that enables query embedding to better match ground truth to improve recall, and constructs a segmentation branch for more precise document image instance segmentation.
- 62
- Recurrent Neural Networks with External Memory for Spoken Language Understanding61
This work proposes to use an external memory to improve memorisation capability of RNNs and demonstrates that the proposed model was able to achieve the state-of-the-art results.
- 58
- Privacy Protection for Open Sharing of Psychiatric and Behavioral Research Data: Ethical Considerations and Recommendations57
This study comprehensively examined the ethical considerations surrounding the acquisition, management, sharing, and utilization of psychiatric and behavioral research data and suggested ethical standards for privacy protection based on varying levels of data sensitivity.
- Learning distributed word representation with multi-contextual mixed embedding57
The proposed MWE model combines the two variants of word2vec in a seamless way via sharing a common encoding structure, which is able to capture the syntax information of words more accurately and incorporates a global text vector into the CBOW variant so as to capture more semantic information.
- MetaNER: Named Entity Recognition with Meta-Learning56
MetaNER incorporates meta-learning and adversarial training strategies to encourage robust, general and transferable representations for sequence labeling and is capable of adapting to new unseen domains with a small amount of annotated data from those domains.
- 55
- Recommendation Algorithm based on Link Prediction and Domain Knowledge in Retail Transactions54
A new recommendation algorithm is proposed, which extends the idea of linkage measure to recommendation in bipartite network, and incorporate domain knowledge with topological property in recommendation process to improve the recommendation quality.
- 53
- Multiscale geographically and temporally weighted regression with a unilateral temporal weighting scheme and its application in the analysis of spatiotemporal characteristics of house prices in Beijing52
An alternative GTWR with a good balance between complexity and interpretability via a unilateral temporal weighting scheme called unilateral GTWR (UGTWR) is proposed, which can effectively capture different levels of spatiotemporal heterogeneity in selected factors affecting house prices at different scales.
- Task assignment strategy for multi-robot based on improved Grey Wolf Optimizer51
The individual speed and position updates in PSO were introduced to enable the gray wolf individual to preserve its optimal location information and accelerate the convergence speed, and the results showed that the IGWO algorithm has faster convergence speed and higher accuracy.
- Causal Hidden Markov Model for Time Series Disease Forecasting50
It is proved that these disease-related hidden variables can be disentangled from others, implying the avoidance of spurious correlation for generalization to medical data from other (out-of-distribution) distributions.
- Movie recommendation based on bridging movie feature and user interest49
A novel hybrid recommendation algorithm is proposed to resolve the two issues by bridging the movie feature and user interest, which aims to make the recommendation results adapt to the change of user interest.
- Encoding Conversation Context for Neural Keyphrase Extraction from Microblog Posts47
A neural keyphrase extraction framework for microblog posts that takes their conversation context into account is presented, where four types of neural encoders, namely, averaged embedding, RNN, attention, and memory networks, are proposed to represent the conversation context.
- A fully automatic computer-aided diagnosis system for hepatocellular carcinoma using convolutional neural networks43
The proposed fully automatic computer-aided diagnosis (CAD) system to diagnose hepatocellular carcinoma (HCC) using convolutional neural network (CNN) classifier is proved satisfactory and feasible by the test set.
- 42
- Joint Learning Embeddings for Chinese Words and their Components via Ladder Structured Networks41
This paper proposes a novel framework, namely, ladder structured networks (LSN), which contains three layers representing word, character and radical and learns their embeddings synchronously and confirms the validity of this approach and shows superiority of the approach over previous work.
- Multi-criteria recommender system based on social relationships and criteria preferences38
The proposed social recommendation model outperforms traditional approaches for both active and cold start users in predicting criteria ratings and confirms that the benefits of nonlinear aggregate functions and cluster analysis, especially when the data set is extremely sparse, are confirmed.
- 38
- 38
- Folding rate prediction based on neural network model38
Back-propagation neural network is adopted to study the relationship between folding rate and protein structure, and it is found that the folding rate depends on CO, LRO and TCD simultaneously, which means that CO and LRO are similarly important in folding rate of protein.
- A Dual-Channel Framework for Sarcasm Recognition by Detecting Sentiment Conflict37
This paper proposes a Dual-Channel Framework by modeling both literal and implied sentiments separately and designs the Dual- Channel Network~(DC-Net) to recognize sentiment conflict.
- Knowledge Overshadowing Causes Amalgamated Hallucination in Large Language Models33
It is shown that the hallucination rate grows with both the imbalance ratio (between the popular and unpopular condition) and the length of dominant condition description, consistent with the derived generalization bound.
- Topic Extraction from Microblog Posts Using Conversation Structures33
This work organize microblog posts as conversation trees based on reposting and replying relations, which enrich context information to alleviate data sparseness and captures the different extents that leader and follower messages may contain the key topical words, thus further enhances the quality of the induced topics.
- Modeling user multiple interests by an improved GCS approach33
The proposed model maps the problem of learning and keeping track of user interests into a clustering and cluster-maintaining problem and provides reasonable performance and high adaptability for learning user multiple interests and their changes.
- Microblog Hashtag Generation via Encoding Conversation Contexts32
This work is the first effort to annotate hashtags with a novel sequence generation framework via viewing the hashtag as a short sequence of words, and proposes to jointly model the target posts and the conversation contexts initiated by them with bidirectional attention.
- What You Say and How You Say it: Joint Modeling of Topics and Discourse in Microblog Conversations32
A neural model to discover word clusters indicating what a conversation concerns and those reflecting how participants voice their opinions are proposed, which can yield both coherent topics and meaningful discourse behavior in microblog conversations.
- Social Media in Citizen-Led Disaster Response: Rescuer Roles, Coordination Challenges, and Untapped Potential.31
This study used semistructured interviews and photo elicitation techniques to better understand how wide-scale rescues occurred during the 2017 Hurricane Harvey flooding in the Greater Houston, Texas USA area and found that citizens used diverse apps and social media-related platforms during these rescues.
- Prompt-Based Code Completion via Multi-Retrieval Augmented Generation30
ProCC, a code completion framework leveraging prompt engineering and the contextual multi-armed bandits algorithm to flexibly incorporate and adapt to multiple perspectives of code, is proposed.
- Microblog Conversation Recommendation via Joint Modeling of Topics and Discourse30
A statistical model is proposed that jointly captures topics for representing user interests and conversation content, and discourse modes for describing user replying behavior and conversation dynamics that outperforms methods that only model content without considering discourse.
- Errors of Machine Translation of terminology in the patent text from English into Chinese28
This paper summarizes eight types of error of terminology in the patent text in the output of Machine Translation from English into Chinese, including term being mistranslated as a verb, term being mistaken for a common noun, and term being Mistranslated due to wrong acronym.
- Trust-Enhanced Cloud Service Selection Model Based on QoS Analysis28
A cloud service selection model based on the trust-enhanced similarity that is able to obtain optimal results via adjusting parameters and exhibits high effectiveness and the cloud services ranking by the model also have better QoS properties than other methods in the comparison experiments.
- 27
- Coupling Global and Local Context for Unsupervised Aspect Extraction26
A novel neural model is proposed, capable of coupling global and local representation to discover aspect words, and it is found that aspect and non-aspect words do exhibit different context, interpreting the superiority of this model in unsupervised aspect extraction.
- Localization and Anderson delocalization of light in fractional dimensions with a quasi-periodic lattice25
Surprisingly, it is found that, in the presence of weak structural disorder, anti-Anderson localization occurs, i.e., the introduced disorder results in an increasing of the size of the linear modes.
- Subtopic-driven Multi-Document Summarization24
A summarization model called STDS is proposed that generates the underlying topic representation from both document view and subtopic view in parallel and outperforms state-of-the-art methods on benchmark datasets.
- Identification of a PTEN mutation with reduced protein stability, phosphatase activity, and nuclear localization in Hong Kong patients with autistic features, neurodevelopmental delays, and macrocephaly24
The I101T mutant has reduced phosphorylation at a PTEN auto‐dephosphorylation site at Threonine 366 and a lowered ratio of nuclear to cytosolic protein level, which may contribute to the tissue overgrowth and autistic features of this PHTS patient.
- Soft Robot Proprioception Using Unified Soft Body Encoding and Recurrent Neural Network22
This work created a synthetic analog to the human muscle system, using paralleled soft pneumatic chambers to serve as receptors for sensing body deformation, and demonstrated the design of three degrees of freedom continuum joint.
- 22
- Biclustering with missing data22
Numerical results show the proposed biclustering method performs well on moderate-sized test cases with even a large missing-value percentage (99%+), and to illustrate the practical usefulness of the method, two case studies are provided.
- New similarity and distance measures of Pythagorean fuzzy sets and its application to selection of advertising platforms21
An approach for multiple criteria group decision making within Pythagorean fuzzy environment is given and another way to express the PFNs and construct a variety of novel distance measures of PFNs and PFSs is given.
- Cross-Media Keyphrase Prediction: A Unified Framework with Multi-Modality Multi-Head Attention and Image Wordings21
This work explores the joint effects of texts and images in predicting the keyphrases for a multimedia post and proposes a novel Multi-Modality Multi-Head Attention (M3H-Att) to capture the intricate cross-media interactions.
- Deep learning time pattern attention mechanism-based short-term load forecasting method20
A novel hybrid deep learning model is proposed for short-term load forecasting that incorporates a time pattern attention mechanism to improve feature extraction and learning capabilities and facilitates a deeper understanding of the impact of external factors on load profiles, leading to the development of a highly accurate load forecasting model.
- Engage the Public: Poll Question Generation for Social Media Posts20
The results show that the model outperforms the popular S2S models without exploiting topics from comments and the dual decoder design can further benefit the prediction of both questions and answers.
- Hashtags, Emotions, and Comments: A Large-Scale Dataset to Understand Fine-Grained Social Emotions to Online Topics20
B baseline performance is examined to predict a topic’s possible social emotions in a multilabel classification setting and results show that a seq2seq model with user comment modeling performs the best, even surpassing human prediction.
- H-INFINITY FILTER BASED PARTICLE FILTER FOR MANEUVERING TARGET TRACKING20
The H1PF alleviates the sample degeneracy problem which is common in the PF, especially when the maneuvers of the target tracking are large and the H1F algorithm can adjust gain imbalance factor by adjusting disturbance attenuation factor and the new algorithm can get the compromise between the accuracy and robustness.
- A Joint Model of Conversational Discourse and Latent Topics on Microblogs19
An unsupervised model that jointly learns word distributions to represent different roles of conversational discourse and various latent topics in reflecting content information can effectively indicate summary-worthy content in microblog conversations.
- Using Content-level Structures for Summarizing Microblog Repost Trees19
A novel repost tree summarization framework is proposed by effectively differentiating two kinds of messages on repost trees called leaders and followers, which are derived from contentlevel structure information, i.e., contents of messages and the reposting relations.
- Mathieu equation with application to analysis of dynamic characteristics of resonant inertial sensors19
Theoretical and experimental analysis show that the correlation of the theoretical curve and the experimental result coincide so perfectly, which means proposed analysis methods for Mathieu equation could be used to analyze the dynamic output characteristic of resonant inertial sensors.
- 17
- 17
- Discrete Argument Representation Learning for Interactive Argument Pair Identification17
Experimental results on the large-scale dataset collected from CMV show that the proposed framework can significantly outperform the competitive baselines and reveal why the model yields superior performance and prove the usefulness of the learned representations.
- What Changed Your Mind: The Roles of Dynamic Topics and Discourse in Argumentation Process17
A novel neural model is proposed that is able to dynamically track the changes of latent topics and discourse in argumentative conversations, allowing the investigation of their roles in influencing the outcomes of persuasion.
- A time-series based aggregation scheme for topic detection in Weibo short texts17
This paper proposes a time-series based aggregation scheme for topic modeling in Weibo that dramatically decreases the topic model perplexity and increase the clustering quality, which also allows for better discovery of underlying topics in weibo.
- 17
- 16
- Property theory of Chinese materia medica: Clinical pharmacodynamics of traditional Chinese medicine15
- Dynamic Online Conversation Recommendation15
This work proposes a neural architecture to exploit changes of user interactions and interests over time, to predict which discussions they are likely to enter and significantly outperforms state-of-the-art models that make a static assumption of user interests.
- Continuity of Topic, Interaction, and Query: Learning to Quote in Online Conversations15
This work studies automatic quotation generation in an online conversation and explores how language consistency affects whether a quotation fits the given context and captures the contextual consistency of a quotation in terms of latent topics, interactions with the dialogue history, and coherence to the query’s existing contents.
- Trending Sentiment-Topic Detection on Twitter15
This work is based on a non-parametric supervised real-time trending topic detection model with sentimental feature that demonstrates impressive effectiveness with 82.3% recall and surpasses all the competitors.
- A Biterm-based Dirichlet Process Topic Model for Short Texts15
This paper proposes a Dirichlet process based on word cooccurrence to make topic mining from short texts more automatically and designs a Markov chain Monte Carlo sampling scheme for posterior inference in the model which is an extension of the sampling algorithm based on Chinese restaurant process.
- 14
- 14
- 13
- Incremental and sequence learning algorithms for weighted regularized extreme learning machines12
The incremental weighted regularized extreme learning machine (IWR-ELM) and the online weighted regularized extreme learning machine (OWR-ELM) are proposed in this paper to enhance WR-ELM’s flexibility and responsiveness by introducing incremental and sequential learning.
- 12
- The application of Deep Reinforcement Learning in Coordinated Control of Nuclear Reactors12
This paper proposes a mathematical model of the coordinated control system, and then transforms it into a reinforcement learning model and develops a deep reinforcement learning control algorithm so-called DDPG algorithm to solve the problem.
- 12
- 12
- Protecting source privacy in federated search12
This paper proposes a two-step federated search protocol that protects the privacy of information sources and is the first attempt to address the research problem of protecting source privacy in federated text search.
- GCS-MA: A group communication system for mobile agents12
This paper proposes a group communication system for mobile agents (GCS-MA), which includes totally ordered multicast and membership management functions, and presents a token-based algorithm, an intra-clusters messaging algorithm and an inter-cluster migration algorithm to achieve atomicity and total ordering properties of multicast messages.
- Diversity in verbal fluency performance and its associations with MRI ‐informed brain age matrices in normal ageing and neurocognitive disorders11
This study aimed to combine the psychometric and morphometric approaches to decode the complex verbal fluency performance in senior adults with normal ageing and neurocognitive disorders.
- Joint Effects of Context and User History for Predicting Online Conversation Re-entries11
A neural framework with three main layers, each modeling context, user history, and interactions between them, is proposed to explore how the conversation context and user chatting history jointly result in their re-entry behavior.
- Improving Distributed Word Representation and Topic Model by Word-Topic Mixture Model11
The experimental results prove that the WTM model performs better on word representation and topic detection compared with some state-of-the-art models.
- Chan–Vese aided fuzzy C‐means approach for whole breast and fibroglandular tissue segmentation: Preliminary application to real‐world breast MRI10
This paper aims to substitute handcrafted breast density segmentation and categorization in MRI with the traditional fuzzy C‐means (FCM) to enable automatic segmentation but may be susceptible to heterogeneity or sparse FGT distribution in MRI.
- Modeling Global and Local Interactions for Online Conversation Recommendation10
A novel framework to automatically recommend conversations to users based on what they said and how they behaved in their chatting histories is presented, which significantly outperforms the state-of-the-art model.
- The Cross-lingual Conversation Summarization Challenge9
A new benchmark is constructed, covering 2 real-world scenarios and 3 language directions, including a low-resource language, that can motivate researches to go beyond English and break the barrier for non-English speakers to benefit from recent advances of conversation summarization.
- 9
- 9
- 9
- 8
- 8
- 8
- Neural Conversation Recommendation with Online Interaction Modeling8
A novel framework to automatically recommend conversations to users based on their prior conversation behaviors is presented, built on neural collaborative filtering and incorporates graph-structured networks to identify salient characteristics from interleaving user interactions.
- 8
- 8
- 8
- 7
- 7
- 7
- 7
- An Efficient Coarse-to-Fine Facet-Aware Unsupervised Summarization Framework based on Semantic Blocks6
This paper proposes an efficient Coarse-to-Fine Facet-Aware Ranking (C2F-FAR) framework for unsupervised long document summarization, which is based on the semantic block, and can achieve new state-of-the-art unsuper supervised summarization results on Gov-Report and BillSum.
- Successful New-entry Prediction for Multi-Party Online Conversations via Latent Topics and Discourse Modeling6
An unsupervised neural network is employed, Variational Auto-Encoder (VAE), to examine the topic content and discourse behavior from newcomer’s chatting history and conversation's ongoing context, and shows that this model significantly outperforms all the baselines and popular neural models.
- Modeling Evolution of Message Interaction for Rumor Resolution6
This paper model confrontation and reciprocity between message pairs via discrete variational autoencoders which effectively reflects the diversified opinion interactivity and captures the variation of message interaction using a hierarchical framework to better integrate information flow of a rumor cascade.
- 6
- 6
- 5
- 5
- When Cantonese NLP Meets Pre-training: Progress and Challenges5
A substantially review of the linguistics and NLP progress for shaping language specificity, resources, and methodologies is reviewed, followed by an introduction to the trendy transformerbased pre-training methods which have been largely advancing the SOTA performance of a wide range of downstream NLP tasks in numerous majority languages.
- 5
- Learning to Rank Microblog Posts for Real-Time Ad-Hoc Search5
This paper combines content-based, microblog-specific and temporal features into learning to rank models, which are found to model microblog posts effectively and are evaluated using tweet data set provided by TERC 2011 and TREC 2012 microblogs.
- 5
- 5
- 5
- BeyondGender: A Multifaceted Bilingual Dataset for Practical Sexism Detection4
BeyondGender is introduced, a dataset meticulously annotated according to the latest definitions of misogyny and misandry that features innovative multifaceted labels encompassing aspects of sexism, gender, phrasing, misogyny, and misandry.
- 4
- 4
- 4
- 4
- 4
- 4
- 4
- 4
- 4
- 4
- 4
- 4
- PlaceField2BVec: A bionic geospatial location encoding method for hierarchical temporal memory model3
- Data-Augmented and Retrieval-Augmented Context Enrichment in Chinese Media Bias Detection3
Based on the previous lexical bias and informational bias structure, this work refine it from the Chinese perspective and goes one step further to craft data with seven fine-grained labels to better understand bias.
- 3
- 3
- 3
- 3
- FedLALR: Client-Specific Adaptive Learning Rates Achieve Linear Speedup for Non-IID Data3
This paper proposes a heterogeneous local variant of AMSGrad, named FedLALR, in which each client adjusts its learning rate based on local historical gradient squares and synchronized learning rates, which enables promising scalability in federated optimization.
- 3
- 3
- 3
- 3
- 3
- 2
- 2
- 2
- 2
- Who Responded to Whom: The Joint Effects of Latent Topics and Discourse in Conversation Structure2
A neural model is proposed to learn latent topics and discourse in word distributions, and predict pairwise initiation-response links via exploiting topic consistency and discourse dependency, which significantly outperforms the previous state of the arts.
- 2
- 2
- 2
- 2
- 2
- 2
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- Suicidal ideation recognition based on sentence completion test via coding- and topic-enhanced model–
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- Two-stage iterative Procrustes match algorithm and its application for VQ-based speaker verification–
- –
- –
- –
- –
- –
- –
- 基于情感时间序列的微博热点主题检测–
A non-parametric supervised real-time-trending topic-detection model with a sentimental feature that detects trending sentimental topic in a short time and demonstrates impressive effectiveness with an 82.3% recall rate, surpassing all of the competitors.
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
Publication data from OpenAlex, with missing venues and authors filled in from Crossref; citation counts are the higher of OpenAlex and Semantic Scholar, last synced 2026-10-16. One-sentence summaries under some papers are written by Semantic Scholar’s model. Citation counts may be lower than on Google Scholar, which indexes more sources.
Report an error
Wrong papers, two people merged into one, or a profile that should not be here? Tell us and we will fix or hide it.