Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileAcademic lineage
Possible advisorsa guess from early papers, not confirmed
- Sheng LiSuggested from co-authorship
Is this you? Claim this profile to confirm or dismiss it.
Works332 from public data
- Generating Natural Language Adversarial Examples through Probability Weighted Word Saliency839
A new word replacement order determined by both the wordsaliency and the classification probability is introduced, and a greedy algorithm called probability weighted word saliency (PWWS) is proposed for text adversarial attack.
- LayoutLMv2: Multi-modal Pre-training for Visually-rich Document Understanding756
LayoutLMv2 architecture with new pre-training tasks to model the interaction among text, layout, and image in a single multi-modal framework and achieves new state-of-the-art results on a wide variety of downstream visually-rich document understanding tasks.
- LTP: A Chinese Language Technology Platform503
LTP (Language Technology Platform) is an integrated Chinese processing platform which includes a suite of high performance natural language processing modules and relevant corpora that achieved good results in some relevant evaluations, such as CoNLL and SemEval.
- Data augmentation approaches in natural language processing: A survey405
This paper frames DA methods into three categories based on the diversity of augmented data, including paraphrasing, noising, and sampling, and introduces their applications in NLP tasks as well as the challenges.
- A Stack-Propagation Framework with Token-Level Intent Detection for Spoken Language Understanding299
A novel framework for SLU to better incorporate the intent information, which further guiding the slot filling is proposed, which achieves the state-of-the-art performance and outperforms other previous methods by a large margin.
- Learning Semantic Hierarchies via Word Embeddings289
This paper proposes a novel and effective method for the construction of semantic hierarchies based on word embeddings, which can be used to measure the semantic relationship between words.
- Recall and Learn: Fine-tuning Deep Pretrained Language Models with Less Forgetting278
A recall and learn mechanism, which adopts the idea of multi-task learning and jointly learns pretraining tasks and downstream tasks, and an Objective Shifting mechanism to focus the learning on downstream tasks gradually.
- Towards Better261
This paper describes the system (HIT-SCIR) submitted to the CoNLL 2018 shared task on Multilingual Parsing from Raw Text to Universal Dependencies, which was ranked first according to LAS and outperformed the other systems by a large margin.
- Towards Conversational Recommendation over Multi-Type Dialogs254
This work creates a human-to-human Chinese dialog dataset DuRecDial, where there are multiple sequential dialogs for a pair of a recommendation seeker and a recommender, where the recommender proactively leads a multi-type dialog to approach recommendation targets and then makes multiple recommendations with rich interaction behavior.
- Few-shot Slot Tagging with Collapsed Dependency Transfer and Label-enhanced Task-adaptive Projection Network210
A Label-enhanced Task-Adaptive Projection Network (L-TapNet) based on the state-of-the-art few-shot classification model – TapNet is proposed, by leveraging label name semantics in representing labels.
- Cross-lingual Dependency Parsing Based on Distributed Representations201
This paper provides two algorithms for inducing cross-lingual distributed representations of words, which map vocabularies from two different languages into a common vector space and bridges the lexical feature gap by using distributed feature representations and their composition.
- CoSDA-ML: Multi-Lingual Code-Switching Data Augmentation for Zero-Shot Cross-Lingual NLP179
A data augmentation framework to generate multi-lingual code-switching data to fine-tune mBERT, which encourages model to align representations from source and multiple target languages once by mixing their context information.
- N-LTP: An Open-source Neural Language Technology Platform for Chinese172
N-LTP, an open-source neural language technology platform supporting six fundamental Chinese NLP tasks: lexical analysis (Chinese word segmentation, part-of-speech tagging, and named entity recognition), syntactic parsing (dependency parsing), and semantic parsing (semantic dependency parsing and semantic role labeling).
- Convolution Neural Network for Relation Extraction172
This paper proposes a novel convolution network, incorporating lexical features, applied to Relation Extraction, and compares the Convolution Neural Network CNN on relation extraction with the state-of-art tree kernel approach, including Typed Dependency Path Kernel and Shortest Dependency path Kernel and Context-Sensitive tree kernel.
- A Co-Interactive Transformer for Joint Slot Filling and Intent Detection156
This paper proposes a co-interactive module to consider the cross-impact by building a bidirectional connection between the two related tasks, where slot and intent can be able to attend on the corresponding mutual information.
- A survey of multilingual large language models148
A unified and thorough review of the field, highlighting recent progress and emerging trends in MLLM research, and provides a unified framework to summarize the current progress in MLLMs.
- Revisiting Embedding Features for Simple Semi-supervised Learning127
Experiments on the task of named entity recognition show that each of the proposed approaches can better utilize the word embedding features, among which the distributional prototype approach performs the best.
- A Survey on Spoken Language Understanding: Recent Advances and New Frontiers126
This paper gives a thorough review of this research, covering different aspects including a new taxonomy, and provides a new perspective for SLU, including single model vs. joint model, implicit joint modeling vs. pretrained paradigm.
- Cross-Lingual BERT Transformation for Zero-Shot Dependency Parsing124
This paper proposes Cross-Lingual BERT Transformation (CLBT), a simple and efficient approach to generate cross-lingual contextualized word embeddings based on publicly available pre-trained BERT models (Devlin et al., 2018).
- AGIF: An Adaptive Graph-Interactive Framework for Joint Multiple Intent Detection and Slot Filling123
An Adaptive Graph-Interactive Framework for joint multiple intent detection and slot filling is proposed, where an intent-slot graph interaction layer is applied to each token adaptively to model the strong correlation between the slot and intents.
- 120
- Named Entity Recognition with Bilingual Constraints108
A method is proposed that formulates the problem of exploring signals on unannotated bilingual text as a simple Integer Linear Program, which encourages entity tags to agree via bilingual constraints and can improve strong baselines for both Chinese and English.
- GL-GIN: Fast and Accurate Non-Autoregressive Model for Joint Multiple Intent Detection and Slot Filling107
This paper proposes a non-autoregressive model for joint multiple intent detection and slot filling, achieving more fast and accurate performance while being 11.5 times faster.
- NL-Augmenter 🦎 → 🐍 A Framework for Task-Sensitive Natural Language Augmentation105
The efficacy of NL-Augmenter is demonstrated by using its transformations to analyze the robustness of popular language models by finding different models to be differently challenged on different tasks, with quasi-systematic score decreases.
- DCR-Net: A Deep Co-Interactive Relation Network for Joint Dialog Act Recognition and Sentiment Classification103
A Deep Co-Interactive Relation Network (DCR-Net) is proposed to explicitly consider the cross-impact and model the interaction between the two tasks by introducing a co-interactive relation layer and employs the Bidirectional Encoder Representation from Transformer in the framework, which can further boost the model's performance in both tasks.
- Joint Extraction of Entities and Relations Based on a Novel Graph Scheme103
This paper converts the joint task into a directed graph by designing a novel graph scheme and proposes a transition-based approach to generate the directed graph incrementally, which can achieve joint learning through joint decoding.
- Towards Better UD Parsing: Deep Contextualized Word Embeddings, Ensemble, and Treebank Concatenation101
- Dynamic Fusion Network for Multi-Domain End-to-end Task-Oriented Dialog99
This work investigates methods that can make explicit use of domain knowledge and introduce a shared-private network to learn shared and specific knowledge and proposes a novel Dynamic Fusion Network (DF-Net) which automatically exploit the relevance between the target domain and each domain.
- A Representation Learning Framework for Multi-Source Transfer Parsing93
This work presents a novel representation learning framework that allows multi-source transfer parsing using full lexical features straightforwardly and significantly outperforms the state-of-the-art transfer system proposed most recently.
- DuRecDial 2.0: A Bilingual Parallel Corpus for Conversational Recommendation88
Experimental results show that the use of additional English data can bring performance improvement for Chinese conversational recommendation, indicating the benefits of DuRecDial 2.0.
- Co-GAT: A Co-Interactive Graph Attention Network for Joint Dialog Act Recognition and Sentiment Classification82
Experimental results on two public datasets show that the proposed Co-Interactive Graph Attention Network successfully captures the two sources of information and achieve the state-of-the-art performance.
- Joint Models for Chinese POS Tagging and Dependency Parsing82
This paper designs several joint models and their corresponding decoding algorithms to incorporate different feature sets and presents an effective pruning strategy to reduce the search space of candidate POS tags, leading to significant improvement of parsing speed.
- Learning Sense-specific Word Embeddings By Exploiting Bilingual Resources79
This paper addresses the problem of representing words with multiple and sense-specific embeddings, which are learned from bilingual parallel data by representing each word with a single embedding, which fails to capture polysemy.
- Large language models meet text-centric multimodal sentiment analysis: a survey77
A comprehensive review of recent research in text-centric multimodal sentiment analysis tasks is presented, the potential of LLMs for text-centric multimodal sentiment analysis is examined, and the challenges and potential research directions for multimodal sentiment analysis in the future are explored.
- Parsing Tweets into Universal Dependencies77
It is shown that it is challenging to deliver consistent annotation due to ambiguity in understanding and explaining tweets and proposed a new method to distill an ensemble of 20 transition-based parsers into a single one that achieves an improvement of 2.2 in LAS over the un-ensembled baseline and outperforms parsers that are state-of-the-art on other treebanks in both accuracy and speed.
- From static to dynamic word representations: a survey76
This survey provides a comprehensive typology of word representation models from a novel perspective that the development from static to dynamic embeddings can effectively address the polysemy problem.
- Character-Level Chinese Dependency Parsing76
This paper presents novel adaptations of two major shift-reduce dependency parsing algorithms to character-level parsing, and demonstrates improved performances over word-based parsing methods on the Chinese Treebank.
- Domain Adaptation for CRF-based Chinese Word Segmentation using Free Annotations73
Experimental results show that the Chinese word segmentation model benefits from free partially annotated data on the SIGHAN Bakeoff 2010 data, and different sources of free annotations are transformed into a unified form of partial annotation.
- Multilingual dependency-based syntactic and semantic parsing72
The authors' CoNLL 2009 Shared Task system includes three cascaded components: syntactic parsing, predicate classification, and semantic role labeling, which wins the first place in the joint task, including both the closed and open challenges.
- Multilingual Large Language Model: A Survey of Resources, Taxonomy and Frontiers70
A thorough review and provide a unified perspective to summarize the recent progress as well as emerging trends in multilingual large language models (MLLMs) literature is presented.
- Knowledge Graph Grounded Goal Planning for Open-Domain Conversation Generation70
Experiments demonstrate that the proposed three-layer Knowledge aware Hierarchical Reinforcement Learning based Model outperforms state of the art baselines in terms of user-interest consistency, dialog coherence, and knowledge accuracy.
- A Neural Transition-Based Approach for Semantic Dependency Graph Parsing69
A neural transition-based parser, using a variant of list-based arc-eager transition algorithm for dependency graph parsing, and two non-trivial improvements are proposed for representing the key components of the transition system, to better capture the semantics of segments and internal sub-graph structures.
- 69
- Few-shot Learning for Multi-label Intent Detection65
The proposed model significantly outperforms strong baselines in both one-shot and five-shot settings and introduces label name embedding as anchor points in representation space, which refines representations of different classes to be well-separated from each other.
- Dynamic Connected Networks for Chinese Spelling Check65
A novel architecture named Dynamic Connected Networks (DCN) is proposed, which generates the candidate Chinese characters via a Pinyin Enhanced Candidate Generator and then utilizes an attention-based network to model the dependencies between two adjacent Chinese characters.
- Entity-Consistent End-to-end Task-Oriented Dialogue System with KB Retriever65
A novel framework which queries the KB in two steps to improve the consistency of generated entities, and introduces a KB retrieval component which explicitly returns the most relevant KB row given a dialogue history.
- 64
- Consistency Regularization for Cross-Lingual Fine-Tuning60
This work uses example consistency regularization to penalize the prediction sensitivity to four types of data augmentations, i.e., subword sampling, Gaussian noise, code-switch substitution, and machine translation, to improve cross-lingual fine-tuning with consistency regularized.
- A Preliminary Evaluation of ChatGPT for Zero-shot Dialogue Understanding58
This work investigates the understanding ability of ChatGPT for zero-shot dialogue understanding tasks including spoken language understanding (SLU) and dialogue state tracking (DST).
- Conversational Graph Grounded Policy Learning for Open-Domain Conversation Generation58
This work presents a novel CG grounded policy learning framework that conducts dialog flow planning by graph traversal, which learns to identify a what-vertex and a how- Vertex from the CG at each turn to guide response generation.
- 58
- Document Modeling with Graph Attention Networks for Multi-grained Machine Reading Comprehension57
A novel multi-grained machine reading comprehension framework that focuses on modeling documents at their hierarchical nature, which are different levels of granularity: documents, paragraphs, sentences, and tokens, which significantly outperforms previous systems at both long and short answer criteria.
- 56
- A Graph-based Method for Entity Linking55
This paper formalizes the task of finding a knowledge base entry that a given named entity mention refers to, namely entity linking, by identifying the most “important” node among the graph nodes representing the candidate entries by introducing three degree-based measures of graph connectivity.
- Joint Word Alignment and Bilingual Named Entity Recognition Using Dual Decomposition52
A graphical model is presented that performs bilingual NER tagging jointly with word alignment, by combining two monolingual tagging models with two unidirectional alignment models, and a dual decomposition inference algorithm is designed to perform joint decoding over the combined alignment and NER output space.
- Transition-Based Disfluency Detection using LSTMs51
This model incrementally constructs and labels the disfluency chunk of input sentences using a new transition system without syntax information, which can capture non-local chunk-level features and is free for noise in syntax.
- Effective Bilingual Constraints for Semi-Supervised Learning of Named Entity Recognizers51
This work proposes a factored probabilistic sequence model that encourages both crosslanguage and intra-document consistency and demonstrates a method that makes effective use of vast amounts of bilingual text to improve monolingual systems.
- Research on Chinese Event Extraction50
An intensive study of the two stages of Chinese event extraction, namely event type recognition and event argument recognition, results in a novel method combining event trigger expansion and a binary classifier while in the step of argument recognition a multi-class classification based on maximum entropy is introduced.
- Cross-Lingual Machine Reading Comprehension48
A novel model called Dual BERT is proposed, which takes advantage of the large-scale training data provided by rich-resource language and learn the semantic relations between the passage and question in bilingual context, and then utilize the learned knowledge to improve reading comprehension performance of low- resource language.
- N-LTP: A Open-source Neural Chinese Language Technology Platform with Pretrained Models.45
N-LTP is introduced, an open-source Python Chinese natural language processing toolkit supporting five basic tasks: Chinese word segmentation, part-of-speech tagging, named entity recognition, dependency parsing, and semantic dependency parsing and is the first toolkit to support all Chinese NLP fundamental tasks.
- A cascaded syntactic and semantic dependency parsing system43
The CoNLL 2008 Shared Task system includes two cascaded components: a syntactic and a semantic dependency parsers, which simplify the traditional four stages semantic role labeling into two: a maximum entropy based argument classification and an ILP-based post inference.
- Chinese Grammatical Error Diagnosis with Long Short-Term Memory Networks42
This paper introduces the Chinese Grammatical Error Diagnosis (CGED) system in the NLP-TEA-3 shared task for CGED, and treats the CGED task as a sequence labeling task and describes three models, including a CRF-based model, an LSTM- based model and an ensemble model using stacking.
- An AMR Aligner Tuned by Transition-based Parser39
A new rich resource enhanced AMR aligner which produces multiple alignments and a new transition system for AMR parsing along with its oracle parser and a transition-based AMR parser that parses a sentence into its AMR graph directly are proposed.
- HIT-SCIR at MRP 2019: A Unified Pipeline for Meaning Representation Parsing via Efficient Training and Effective Encoding38
Generally, this paper proposed a unified pipeline to meaning representation parsing, including framework-specific transition-based parsers, BERT-enhanced word representation, and post-processing, and especially ranked first in UCCA framework.
- Transition-Based Syntactic Linearization37
This work extends the transition system so that word ordering is performed in addition to syntactic parsing, resulting in a linearization system that runs in guaranteed quadratic time and runs an order of magnitude faster than a state-of-the-art baseline using best-first search.
- 37
- 36
- Understanding Medical Conversations with Scattered Keyword Attention and Weak Supervision from Responses36
This work approaches the medical slot filling as a multi-label classification problem with label-embedding attentive model to pay more attention to scattered medical keywords and learn the classification models by weak-supervision from responses.
- 35
- 32
- 32
- A Grammar-driven Convolution Tree Kernel for Semantic Role Classification32
A grammardriven convolution tree kernel for semantic role classification by introducing more linguistic knowledge into the standard tree kernel is proposed and a composite kernel to integrate feature-based and tree kernel-based methods is presented.
- 31
- Slot-consistent NLG for Task-oriented Dialogue Systems with Iterative Rectification Network30
Iterative Rectification Network (IRN) is proposed for improving general NLG systems to produce both correct and fluent responses and applies a bootstrapping algorithm to sample training candidates and uses reinforcement learning to incorporate discrete reward related to slot inconsistency into training.
- 30
- 29
- 29
- 28
- ViTCoT: Video-Text Interleaved Chain-of-Thought for Boosting Video Understanding in Large Language Models26
This work introduces a novel video reasoning paradigm: Video-Text Interleaved CoT (ViTCoT), which facilitates more intuitive and cognitively aligned reasoning and extensively explores the potential of the ViTCoT paradigm in the video understanding field.
- 26
- 26
- Towards reasoning era: a survey of long chain-of-thought for reasoning large language models25
This survey distinguishes Long CoT from Short CoT and introduces a novel taxonomy to categorize current reasoning paradigms, and explores the key characteristics of Long CoT: deep reasoning, extensive exploration, and feasible reflection, which enable models to handle more complex tasks and produce more efficient, coherent outcomes compared to the shallower Short CoT.
- 25
- 24
- 24
- 24
- 24
- 24
- 23
- Don’t be Contradicted with Anything! CI-ToD: Towards Benchmarking Consistency for Task-oriented Dialogue System23
This paper introduces CI-ToD, a novel dataset for Consistency Identification in Task-oriented Dialog system, and annotates the single label to enable the model to judge whether the system response is contradictory, but also provides more fine-grained labels to encourage model to know what inconsistent sources lead to it.
- 23
- 23
- 22
- 22
- Injecting Word Information with Multi-Level Word Adapter for Chinese Spoken Language Understanding22
Experimental results on two Chinese SLU datasets show that the proposed multi-level word adapter can capture useful word information and achieve state-of-the-art performance.
- Distilling Knowledge for Search-based Structured Prediction22
Experimental results show that distillation can effectively improve the single model’s performance and the final model achieves improvements of 1.32 in LAS and 2.65 in BLEU score on these two tasks respectively and it outperforms the greedy structured prediction models in previous literatures.
- 21
- 20
- 20
- 20
- 20
- 19
- 19
- 19
- 19
- 19
- 19
- 19
- 18
- 18
- 18
- 17
- What are the Essential Factors in Crafting Effective Long Context Multi-Hop Instruction Datasets? Insights and Best Practices16
This work proposes the Multi-agent Interactive Multi-hop Generation (MIMG) framework, incorporating a Quality Verification Agent, a Single-hop Question Generation Agent, a Multiple Question Sampling Strategy, and a Multi-hop Question Merger Agent that improves the data quality.
- 16
- Using a Hybrid Convolution Tree Kernel for Semantic Role Labeling16
A composite kernel is presented through combining the hybrid convolution tree kernel method with a feature-based method extended by the polynomial kernel, which achieves better performance than each of the individual methods and outperforms the best reported system on the CoNLL-2005 corpus.
- 15
- 15
- 15
- Multi-domain Spoken Language Understanding Using Domain- and Task-aware Parameterization15
This article proposes to improve the parameterization of this method by using domain-specific and task-specific model parameters for fine-grained knowledge representation and transfer and shows its transferability when adapting to a new domain with little data, outperforming the prior best model.
- 15
- 15
- 14
- 14
- 14
- 14
- 14
- Can Large Language Models Understand You Better? An MBTI Personality Detection Dataset Aligned with Population Traits13
Experimental results not only highlight the polarized predictions and biases in LLMs as key directions for future research, but also confirm that soft labels can provide more benefits to other psychological tasks than hard labels.
- Python is Not Always the Best Choice: Embracing Multilingual Program of Thoughts13
This paper proposes a task and model agnostic approach called MultiPoT, which achieves comparable or superior performance compared to the best monolingual PoT in almost all tasks across all models and significantly outperforms Python Self-Consistency.
- 13
- 13
- 13
- 13
- 13
- Appraisal Expression Recognition Based on Syntactic Path13
Experimental results on the camera and MP3 player domains show that syntactic paths are helpful for appraisal expression recognition, and an edit distance based syntactic path matching method can further improve the performance of appraisal expressions recognition.
- Divide-Solve-Combine: An Interpretable and Accurate Prompting Framework for Zero-shot Multi-Intent Detection12
By decomposition, DSCP allows the model to track the explicit reasoning process and improve the interpretability, and proposes an interactive divide-solve-combine prompting (Inter-DSCP) to naturally capture the interaction capabilities of large language models.
- 12
- 12
- 12
- 12
- Deep Contextualized Word Embeddings for Universal Dependency Parsing12
Based on ELMo’s advantage on OOV, experiments that simulate low-resource settings are conducted and the results show that deep contextualized word embeddings are effective for data-insufficient tasks where the OOV problem is severe.
- Sentence Compression for Target-Polarity Word Collocation Extraction12
This work applies a discriminative conditional random field model, with some special sentimentrelated features, in order to automatically compress sentiment sentences.
- 12
- Text Is No More Enough! A Benchmark for Profile-Based Spoken Language Understanding11
This paper introduces a new and important task, Profile-based Spoken Language Understanding (ProSLU), which requires the model that not only relies on the plain text but also the supporting profile information to predict the correct intents and slots, and introduces a large-scale human-annotated Chinese dataset with over 5K utterances.
- 11
- 10
- 10
- 10
- 10
- 10
- 10
- 10
- 10
- 10
- 8
- 8
- 8
- 7
- Towards Complex Scenarios: Building End-to-End Task-Oriented Dialogue System across Multiple Knowledge Bases7
This work introduces a KB-over-KB Heterogeneous Graph Attention Network (KoK-HAN) to facilitate model to reason over multiple KBs and empirically shows that the existing single-KB EToDs fail to work on multi-KB settings that require models to reason across various KBs.
- 7
- 7
- 7
- 7
- 7
- Deep Learning in Lexical Analysis and Parsing7
In this tutorial, a review of each line of work is given, by contrasting them with traditional statistical methods, and organizing them in consistent orders.
- 7
- RBF++: Quantifying and Optimizing Reasoning Boundaries Across Measurable and Unmeasurable Capabilities for Chain-of-Thought Reasoning6
This work defines the reasoning boundary (RB) as the maximum limit of CoT performance, and proposes a combination law for RBs, enabling quantitative analysis and offering actionable guidance across various CoT tasks, and introduces a constant assumption, which replaces unmeasurable RBs with scenario-specific constants.
- CroPrompt: Cross-task Interactive Prompting for Zero-shot Spoken Language Understanding6
This work presents the pioneering work of Cross-task Interactive Prompting (CroPrompt) for SLU, which enables the model to interactively leverage the information exchange across the correlated tasks in SLU and introduces a multi-task self-consistency mechanism to mitigate the error propagation caused by the intent information injection.
- 6
- 6
- 6
- 6
- 6
- 6
- 6
- 6
- 6
- 6
- Learning the Boundary of Solvability: Aligning LLMs to Detect Unsolvable Problems5
A multi-domain dataset containing both solvable and unsolvable questions, UnsolvableQA, together with an alignment framework, UnsolvableRL, and a reinforcement learning paradigm that balances objective unsolvability detection with calibrated confidence under capability limits are proposed.
- 5
- 5
- 5
- 5
- 5
- 5
- TD-GIN: Token-level Dynamic Graph-Interactive Network for Joint Multiple Intent Detection and Slot Filling5
This paper proposes a Token-level Dynamic Graph-Interactive Network (TD-GIN) for joint multiple intent detection and slot filling, where it model the interaction between multiple intents and each token slot in a unified graph architecture.
- 5
- 5
- 5
- Jointly or Separately: Which is Better for Parsing Heterogeneous Dependencies?5
This paper proposes a joint model which can make full use of the correlations between heterogeneous dependencies, and can answer the following question: parsing heterogeneous Dependency Parsing jointly or separately, which is better?
- 5
- 4
- Improving Consistency Identification in Task-oriented Dialogue Through Multi-Agent Collaboration4
A multi-agent collaboration framework (MAC-CIToD) is introduced to model the interaction across three sub-tasks in CI-ToD, including (1) Full Connection paradigm, (2) Cycle Connection paradigm, and (3) Central Connection paradigm, which effectively builds interaction across QI, HI, and KBI.
- 4
- 4
- 4
- 4
- Interactive Gated Decoder for Machine Reading Comprehension4
An Interactive Gated Decoder (IG Decoder), which focuses on modeling the interactions between the answer span prediction and no-answer prediction with a gating mechanism, and a simple but effective approach for automatically generating pseudo training data, which aims to enrich the training data of the unanswerable questions.
- 4
- Stacking Heterogeneous Joint Models of Chinese POS Tagging and Dependency Parsing4
A constituent-based joint model which derives the POS tag sequence and dependency tree from the output of PCFG parsers, and then integrate it into the guided transition- based joint model achieves the best performance.
- 4
- 3
- Beyond Surface Reasoning: Unveiling the True Long Chain-of-Thought Capacity of Diffusion Large Language Models3
Several practical mitigations are proposed, including parallel-oriented prompting, diffusion early stopping, and parallel scaling, to reduce PSC-induced ineffectiveness and inefficiencies.
- 3
- MPCC: A Novel Benchmark for Multimodal Planning with Complex Constraints in Multimodal Large Language Models3
This work formalizes multimodal constraints in planning, provides a rigorous evaluation framework, and highlights the need for advancements in constraint-aware reasoning for real-world MLLM applications.
- 3
- 3
- 3
- 3
- Pro-HAN: A Heterogeneous Graph Attention Network for Profile-based Spoken Language Understanding3
This work introduces a Heterogeneous Graph Attention Network to perform reasoning across multiple Profile information, called Pro-HAN, and establishes a new state-of-the-art on the ProSLU dataset, with an improvement of approximately 8% across all three metrics.
- 3
- 3
- 3
- 3
- 3
- 3
- A Semantics Oriented Grammar for Chinese Treebanking3
This paper introduces a semantics oriented grammar for Chinese, designed to provide basic supports for tasks such as automatic semantic parsing and sentence generation, and is the first Chinese grammar representation capable of direct transformation into logical forms.
- Manager: Aggregating Insights From Unimodal Experts in Two-Tower VLMs and MLLMs2
Manager is proposed, a lightweight, efficient and effective plugin that adaptively aggregates insights from different levels of pre-trained unimodal experts to facilitate more comprehensive VL alignment and fusion and significantly boosts the zero-shot performance of LLaVA-OV across different categories of capabilities, images, and resolutions on 20 downstream datasets, whether the multi-grid algorithm is enabled or not.
- 2
- 2
- 2
- 2
- 2
- 2
- Exploring Multi-Grained Concept Annotations for Multimodal Large Language Models2
The potential of MMGiC to help MLLMs better locate and learn concepts, aligning vision and language at multiple granularities is clearly explored and demonstrated.
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- Multiple TreeBanks Integration for Chinese Phrase Structure Grammar Parsing Using Bagging2
Bagging is proposed to enhance the baseline performance of PSG parsing in CIPS-Bakeoff2012 Task3.0 and outperforms the best closed system by nearly 4% which uses a single model.
- Micro blogs Oriented Word Segmentation System2
A Chinese word segmentation system submitted to the first task on CLP 2012 back-offs is presented using a conditional random field sequence model and derives a micro blog specified lexicon from auto-analyzed data and use lexicon related features to assist the model.
- 2
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- MPFToD: a modularized pre-training framework for consistency identification in task-oriented dialogue1
This paper introduces a modularized pre-training framework (MPFToD), which is capable of utilizing large amounts of KB-free dialogues and proposes three pre-training tasks including query response matching pre-training; dialogue history consistent identification pre-training; and KB mask language modeling to enhance different abilities of CI-ToD model.
- 1
- 1
- 1
- 1
- Beyond Static Evaluation: A Dynamic Approach to Assessing AI Assistants’ API Invocation Capabilities1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- End-to-End Task-Oriented Dialogue System with Distantly Supervised Knowledge Base Retriever1
This paper introduces an explicit KB retrieval component (KB retriever) into the seq2seq dialogue system and significantly outperforms other sequence-to-sequence based baseline models on both automatic and human evaluation.
- 1
- 1
- 1
- Fast Computing Grammar-driven Convolution Tree Kernel for Semantic Role Labeling.1
Two fast grammar-driven convolution tree kernel algorithms are designed, which can compute the GTK in polynomial time, and Experimental results on the CoNLL-2005 SRL data show that the two FGTK algorithms are much faster than theGTK.
- 1
- –
- –
- –
- –
- –
- –
- –
- –
- –
- Learning-to-Context Slope: Evaluating In-Context Learning Effectiveness Beyond Performance Illusions–
- –
- –
- –
- –
- –
- –
- –
- –
- ESAinsTOD: a unified end-to-end schema-aware instruction-tuning framework for task-oriented dialog modeling–
ESAinsTOD, a unified End-to-end Schema-Aware Instruction-tuning framework for general Task-Oriented Dialog modeling, introduces a structured methodology to go beyond simply fine-tuning Large Language Models, enabling flexible adaptation to various dialogue task flows and schemas.
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- HC-search for incremental parsing–
Inspired by recently proposed HC-search framework, this work decomposes the incremental parsing algorithm into two steps: first searching a set of high-quality outputs with beam-search, and second selecting the best output with a ranking model.
- –
- –
- –
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 2026-10-13. 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.