Author name cluster

Ke Ji

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

19 papers

2 author rows

EAAI Journal 2026 Journal Article

ESEN: Evidence-aware Semantic Enhancement Network for Fact-checking Fake News Detection

Yanfang Qiu
Kun Ma
Xiaoyun Liu
Ke Ji
Zhenxiang Chen
Bo Yang

Fact-checking fake news detection is a challenging task that requires retrieving multiple pieces of evidence to verify the truthfulness of a claim. Despite the effectiveness of current methods, they still suffer from the following problems: (1) neglecting attentional constraints on important words related to the claim; (2) insufficiently exploring the relationships between different pieces of evidence. To address these issues, we propose an Evidence-aware Semantic Enhancement Network for Fact-checking Fake News Detection (abbreviated as ESEN). First, we model the claim and evidence as a graph structure to extract semantic information, putting the attentive guidance to closely follow the syntactic information extraction. Then, we incorporate the extracted syntactic information into the semantic information. Finally, we propose an information interaction network for evidence perception to capture multi-level interactive information and enhance the interaction between evidences by constructing three types attention layer. Evaluation experiments are performed on the Snopes and PolitiFact datasets. Experimental results and ablation studies show that our ESEN outperforms other baseline models. Our source code is available at https: //github. com/makunjida/ESEN.

Details DOI

NeurIPS Conference 2025 Conference Paper

QFFT, Question-Free Fine-Tuning for Adaptive Reasoning

Wanlong Liu
Junxiao Xu
Fei Yu
Yukang Lin
Ke Ji
Wenyu Chen
Lifeng Shang
Yasheng Wang

Recent advancements in Long Chain-of-Thought (CoT) reasoning models have improved performance on complex tasks, but they suffer from overthinking, which generates redundant reasoning steps, especially for simple questions. This paper revisits the reasoning patterns of Long and Short CoT models, observing that the Short CoT patterns offer concise reasoning efficiently, while the Long CoT patterns excel in challenging scenarios where the Short CoT patterns struggle. To enable models to leverage both patterns, we propose Question-Free Fine-Tuning (QFFT), a fine-tuning approach that removes the input question during training and learns exclusively from Long CoT responses. This approach enables the model to adaptively employ both reasoning patterns: it prioritizes the Short CoT patterns and activates the Long CoT patterns only when necessary. Experiments on various mathematical datasets demonstrate that QFFT reduces average response length by more than 50\%, while achieving performance comparable to Supervised Fine-Tuning (SFT). Additionally, QFFT exhibits superior performance compared to SFT in noisy, out-of-domain, and low-resource scenarios.

PDF Details

ICLR Conference 2025 Conference Paper

SmartRAG: Jointly Learn RAG-Related Tasks From the Environment Feedback

Jingsheng Gao
Linxu Li
Ke Ji
Weiyuan Li
Yixin Lian
Yuzhuo Fu
Bin Dai

RAG systems consist of multiple modules to work together. However, these modules are usually separately trained. We argue that a system like RAG that incorporates multiple modules should be jointly optimized to achieve optimal performance. To demonstrate this, we design a specific pipeline called SmartRAG that includes a policy network and a retriever. The policy network can serve as 1) a decision maker that decides when to retrieve, 2) a query rewriter to generate a query most suited to the retriever and 3) an answer generator that produces the final response with/without the observations. We then propose to jointly optimize the whole system using a reinforcement learning algorithm, with the reward designed to encourage the system to achieve the highest performance with minimal retrieval cost. When jointly optimized, each module can be aware of how other modules are working and thus find the best way to work together as a complete system. Empirical results demonstrate that the jointly optimized system can achieve better performance than separately optimized counterparts.

Details

NeurIPS Conference 2025 Conference Paper

The First Few Tokens Are All You Need: An Efficient and Effective Unsupervised Prefix Fine-Tuning Method for Reasoning Models

Ke Ji
Jiahao Xu
Tian Liang
Qiuzhi Liu
Zhiwei He
Xiaoyuan Liu
Xingyu Chen
Junying Chen

Improving the reasoning capabilities of large language models (LLMs) typically requires supervised fine-tuning with labeled data or computationally expensive sampling. We introduce Unsupervised Prefix Fine-Tuning (UPFT), which leverages the observation of Prefix Self-Consistency -- the shared initial reasoning steps across diverse solution trajectories -- to enhance LLM reasoning efficiency. By training exclusively on the initial prefix substrings (as few as 8 tokens), UPFT removes the need for labeled data or exhaustive sampling. Experiments on reasoning benchmarks show that UPFT matches the performance of supervised methods such as Rejection Sampling Fine-Tuning, while reducing training time by 75\% and sampling cost by 99\%. Further analysis reveals that errors tend to appear in later stages of the reasoning process and that prefix-based training preserves the model’s structural knowledge. This work demonstrates how minimal unsupervised fine-tuning can unlock substantial reasoning gains in LLMs, offering a scalable and resource-efficient alternative to conventional approaches.

PDF Details

EAAI Journal 2024 Journal Article

DIMN: Dual Integrated Matching Network for multi-choice reading comprehension

Qiang Wei
Kun Ma
Xinyu Liu
Ke Ji
Bo Yang
Ajith Abraham

Multi-choice reading comprehension is a task that involves selecting the correct option from a set of option choices. Recently, the attention mechanism has been widely used to acquire embedding representations. However, there are two significant challenges: (1) generating the contextualized representations, namely, drawing associated information, and (2) capturing the global interactive relationship, namely, drawing local semantics. To address these issues, we have proposed the Dual Integrated Matching Network (DIMN) for multi-choice reading comprehension. It consists of two major parts. Fusing Information from Passage and Question-option pair into Enhanced Embedding Representation (FEER) is proposed to draw associated information to enhance embedding representation, which incorporates the information that reflects the most salient supporting entities to answer the question into the contextualized representations; Linear Integration of Co-Attention and Convolution (LIAC) is proposed to capture the interactive information and local semantics to construct global interactive relationship, which incorporates local semantics of a single sequence into the question-option-aware passage and passage-aware question-option representation. The experiments are shown that our DIMN performs better accuracy on three datasets: RACE (69. 34%), DREAM (68. 45%) and MCTest (71. 81% on MCTest160 and 78. 83% on MCTest500). Our DIMN is beneficial for improving the ability of machines to understand natural language. The system we have developed has been applied to customer service support. Our source code is accessible at https: //github. com/vqiangv/DIMN.

Details DOI

IJCAI Conference 2024 Conference Paper

Domain-Hierarchy Adaptation via Chain of Iterative Reasoning for Few-shot Hierarchical Text Classification

Ke Ji
Peng Wang
Wenjun Ke
Guozheng Li
Jiajun Liu
Jingsheng Gao
Ziyu Shang

Recently, various pre-trained language models (PLMs) have been proposed to prove their impressive performances on a wide range of few-shot tasks. However, limited by the unstructured prior knowledge in PLMs, it is difficult to maintain consistent performance on complex hierarchically dependent tasks, especially when the downstream data is extremely scarce. The main challenge is how to transfer the unstructured semantic space in PLMs to the downstream domain hierarchy. Unlike previous work on hierarchical text classification (HTC) which directly performs multi-label classification or uses graph neural network (GNN) to inject label hierarchy, in this work, we study the HTC problem under a few-shot setting to adapt knowledge in PLMs from an unstructured manner to the downstream hierarchy. Technically, we design a simple yet effective method named Hierarchical Iterative Conditional Random Field (HierICRF) to search the most domain-challenging directions and exquisitely crafts domain-hierarchy adaptation as a hierarchical iterative language modeling problem, and then it encourages the model to make hierarchical consistency self-correction during the inference, thereby achieving knowledge transfer with hierarchical consistency preservation. We perform HierICRF on various architectures, and extensive experiments on two popular HTC datasets demonstrate that prompt with HierICRF significantly boosts the few-shot HTC performance with an average Micro-F1 by 28. 80% to 1. 50% and Macro-F1 by 36. 29% to 1. 5% over the previous state-of-the-art (SOTA) baselines under few-shot settings (1->16), while remaining SOTA hierarchical consistency performance.

PDF Details DOI

IJCAI Conference 2024 Conference Paper

Empirical Analysis of Dialogue Relation Extraction with Large Language Models

Guozheng Li
Zijie Xu
Ziyu Shang
Jiajun Liu
Ke Ji
Yikai Guo

Dialogue relation extraction (DRE) aims to extract relations between two arguments within a dialogue, which is more challenging than standard RE due to the higher person pronoun frequency and lower information density in dialogues. However, existing DRE methods still suffer from two serious issues: (1) hard to capture long and sparse multi-turn information, and (2) struggle to extract golden relations based on partial dialogues, which motivates us to discover more effective methods that can alleviate the above issues. We notice that the rise of large language models (LLMs) has sparked considerable interest in evaluating their performance across diverse tasks. To this end, we initially investigate the capabilities of different LLMs in DRE, considering both proprietary models and open-source models. Interestingly, we discover that LLMs significantly alleviate two issues in existing DRE methods. Generally, we have following findings: (1) scaling up model size substantially boosts the overall DRE performance and achieves exceptional results, tackling the difficulty of capturing long and sparse multi-turn information; (2) LLMs encounter with much smaller performance drop from entire dialogue setting to partial dialogue setting compared to existing methods; (3) LLMs deliver competitive or superior performances under both full-shot and few-shot settings compared to current state-of-the-art; (4) LLMs show modest performances on inverse relations but much stronger improvements on general relations, and they can handle dialogues of various lengths especially for longer sequences.

PDF Details DOI

IJCAI Conference 2024 Conference Paper

Fast and Continual Knowledge Graph Embedding via Incremental LoRA

Jiajun Liu
Wenjun Ke
Peng Wang
Jiahao Wang
Jinhua Gao
Ziyu Shang
Guozheng Li
Zijie Xu

Continual Knowledge Graph Embedding (CKGE) aims to efficiently learn new knowledge and simultaneously preserve old knowledge. Dominant approaches primarily focus on alleviating catastrophic forgetting of old knowledge but neglect efficient learning for the emergence of new knowledge. However, in real-world scenarios, knowledge graphs (KGs) are continuously growing, which brings a significant challenge to fine-tuning KGE models efficiently. To address this issue, we propose a fast CKGE framework (FastKGE), incorporating an incremental low-rank adapter (IncLoRA) mechanism to efficiently acquire new knowledge while preserving old knowledge. Specifically, to mitigate catastrophic forgetting, FastKGE isolates and allocates new knowledge to specific layers based on the fine-grained influence between old and new KGs. Subsequently, to accelerate fine-tuning, FastKGE devises an efficient IncLoRA mechanism, which embeds the specific layers into incremental low-rank adapters with fewer training parameters. Moreover, IncLoRA introduces adaptive rank allocation, which makes the LoRA aware of the importance of entities and adjusts its rank scale adaptively. We conduct experiments on four public datasets and two new datasets with a larger initial scale. Experimental results demonstrate that FastKGE can reduce training time by 34%-49% while still achieving competitive link prediction performance against state-of-the-art models on four public datasets (average MRR score of 21. 0% vs. 21. 1%). Meanwhile, on two newly constructed datasets, FastKGE saves 51%-68% training time and improves link prediction performance by 1. 5%.

PDF Details DOI

EAAI Journal 2024 Journal Article

G-HFIN: Graph-based Hierarchical Feature Integration Network for propaganda detection of We-media news articles

Xinyu Liu
Kun Ma
Qiang Wei
Ke Ji
Bo Yang
Ajith Abraham

In the era of We-media, articles are written by independently individuals that are not officially registered with the authorities. Propaganda hidden in the We-media articles have the potential to polarize public opinion and influence the mindset of the target audience. Currently, graph neural networks (GNNs) have been remarkably successful in Natural Language Processing (NLP). However, there are still some challenges to apply existing GNN-based for propaganda detection due to the limitation of extracting diverse word dependencies and capturing non-consecutive and long-range context. In this paper, we have proposed a Graph-based Hierarchical Feature Integration Network (abbreviated as G-HFIN) for Propaganda Detection. Specifically, semantic, syntactic, and sequential features are extracted to construct three heterogeneous graphs. Then, the Residual-connected Dual-layer Coarsening and Refining procedures (abbreviated as RDCR) are proposed to promote information interactions between distant nodes that are not directly connected, preserving both local and global node information during the intra-graph information propagation. Subsequently, an Attention-based Three-channel Feature Integration (abbreviated as ATFI) is proposed to harmonize sequence, semantic and syntactic information from three channels during the inter-graph information enhancement. Intra-graph and Inter-graph Joint Information Propagation is to implement homogeneous and heterogeneous information interaction respectively. Finally, these news representations are pooled and fed into the propaganda detection classifier. The experiments on three public datasets demonstrate that our model has outperformed state-of-the-art methods.

Details DOI

IJCAI Conference 2024 Conference Paper

Incorporating Schema-Aware Description into Document-Level Event Extraction

Zijie Xu
Peng Wang
Wenjun Ke
Guozheng Li
Jiajun Liu
Ke Ji
Xiye Chen
Chenxiao Wu

Document-level event extraction (DEE) aims to extract the structured event information from a given document, facing two critical challenges: (1) event arguments always scatter across sentences (arguments-scattering); (2) multiple events can co-occur in one document (multi-event). Most recent studies mainly follow two simplified settings to ease the challenges: one simplifies DEE with the no-trigger-words design (NDEE), and the other focuses on event argument extraction (DEAE), a sub-task of DEE. However, the former excludes trigger extraction and suffers from error propagation in the sub-tasks. The latter relies heavily on the gold triggers as prerequisites and struggles to distinguish multiple arguments playing the same role in different events. To address the limitations above, we propose a novel joint trigger and argument extraction paradigm SEELE to enhance the DEE model via incorporating SchEma-awarE descriptions into Document-Level Event extraction. Specifically, the schema-aware descriptions are leveraged from two aspects: (1) guiding the attention mechanism among event-aware tokens across sentences, which relieves arguments-scattering without error propagation; (2) performing the fine-grained contrastive learning to distinguish different events, which mitigates multi-event without gold triggers. Extensive experiments show the superiority of SEELE, achieving notable improvements (2. 1% to 9. 7% F1) on three NDEE datasets and competitive performance on two DEAE datasets. Our code is available at https: //github. com/TheoryRhapsody/SEELE.

PDF Details DOI

AAAI Conference 2024 Conference Paper

LAMM: Label Alignment for Multi-Modal Prompt Learning

Jingsheng Gao
Jiacheng Ruan
Suncheng Xiang
Zefang Yu
Ke Ji
Mingye Xie
Ting Liu
Yuzhuo Fu

With the success of pre-trained visual-language (VL) models such as CLIP in visual representation tasks, transferring pre-trained models to downstream tasks has become a crucial paradigm. Recently, the prompt tuning paradigm, which draws inspiration from natural language processing (NLP), has made significant progress in VL field. However, preceding methods mainly focus on constructing prompt templates for text and visual inputs, neglecting the gap in class label representations between the VL models and downstream tasks. To address this challenge, we introduce an innovative label alignment method named \textbf{LAMM}, which can dynamically adjust the category embeddings of downstream datasets through end-to-end training. Moreover, to achieve a more appropriate label distribution, we propose a hierarchical loss, encompassing the alignment of the parameter space, feature space, and logits space. We conduct experiments on 11 downstream vision datasets and demonstrate that our method significantly improves the performance of existing multi-modal prompt learning models in few-shot scenarios, exhibiting an average accuracy improvement of 2.31(\%) compared to the state-of-the-art methods on 16 shots. Moreover, our methodology exhibits the preeminence in continual learning compared to other prompt tuning methods. Importantly, our method is synergistic with existing prompt tuning methods and can boost the performance on top of them. Our code and dataset will be publicly available at https://github.com/gaojingsheng/LAMM.

PDF Details DOI

IJCAI Conference 2024 Conference Paper

Meta In-Context Learning Makes Large Language Models Better Zero and Few-Shot Relation Extractors

Guozheng Li
Peng Wang
Jiajun Liu
Yikai Guo
Ke Ji
Ziyu Shang
Zijie Xu

Relation extraction (RE) is an important task that aims to identify the relationships between entities in texts. While large language models (LLMs) have revealed remarkable in-context learning (ICL) capability for general zero and few-shot learning, recent studies indicate that current LLMs still struggle with zero and few-shot RE. Previous studies are mainly dedicated to design prompt formats and select good examples for improving ICL-based RE. Although both factors are vital for ICL, if one can fundamentally boost the ICL capability of LLMs in RE, the zero and few-shot RE performance via ICL would be significantly improved. To this end, we introduce Micre (Meta In-Context learning of LLMs for Relation Extraction), a new meta-training framework for zero and few-shot RE where an LLM is tuned to do ICL on a diverse collection of RE datasets (i. e. , learning to learn in context for RE). Through meta-training, the model becomes more effectively to learn a new RE task in context by conditioning on a few training examples with no parameter updates or task-specific templates at inference time, enabling better zero and few-shot task generalization. We experiment Micre on various LLMs with different model scales and 12 public RE datasets, and then evaluate it on unseen RE benchmarks under zero and few-shot settings. Micre delivers comparable or superior performance compared to a range of baselines including supervised fine-tuning and typical in-context learning methods. We find that the gains are particular significant for larger model scales, and using a diverse set of the meta-training RE datasets is key to improvements. Empirically, we show that Micre can transfer the relation semantic knowledge via relation label name during inference on target RE datasets.

PDF Details DOI

AAAI Conference 2024 Conference Paper

OntoFact: Unveiling Fantastic Fact-Skeleton of LLMs via Ontology-Driven Reinforcement Learning

Ziyu Shang
Wenjun Ke
Nana Xiu
Peng Wang
Jiajun Liu
Yanhui Li
Zhizhao Luo
Ke Ji

Large language models (LLMs) have demonstrated impressive proficiency in information retrieval, while they are prone to generating incorrect responses that conflict with reality, a phenomenon known as intrinsic hallucination. The critical challenge lies in the unclear and unreliable fact distribution within LLMs trained on vast amounts of data. The prevalent approach frames the factual detection task as a question-answering paradigm, where the LLMs are asked about factual knowledge and examined for correctness. However, existing studies primarily focused on deriving test cases only from several specific domains, such as movies and sports, limiting the comprehensive observation of missing knowledge and the analysis of unexpected hallucinations. To address this issue, we propose OntoFact, an adaptive framework for detecting unknown facts of LLMs, devoted to mining the ontology-level skeleton of the missing knowledge. Specifically, we argue that LLMs could expose the ontology-based similarity among missing facts and introduce five representative knowledge graphs (KGs) as benchmarks. We further devise a sophisticated ontology-driven reinforcement learning (ORL) mechanism to produce error-prone test cases with specific entities and relations automatically. The ORL mechanism rewards the KGs for navigating toward a feasible direction for unveiling factual errors. Moreover, empirical efforts demonstrate that dominant LLMs are biased towards answering Yes rather than No, regardless of whether this knowledge is included. To mitigate the overconfidence of LLMs, we leverage a hallucination-free detection (HFD) strategy to tackle unfair comparisons between baselines, thereby boosting the result robustness. Experimental results on 5 datasets, using 32 representative LLMs, reveal a general lack of fact in current LLMs. Notably, ChatGPT exhibits fact error rates of 51.6% on DBpedia and 64.7% on YAGO, respectively. Additionally, the ORL mechanism demonstrates promising error prediction scores, with F1 scores ranging from 70% to 90% across most LLMs. Compared to the exhaustive testing, ORL achieves an average recall of 80% while reducing evaluation time by 35.29% to 63.12%.

PDF Details DOI

EAAI Journal 2024 Journal Article

Random color transformation for single domain generalized retinal image segmentation

Song Guo
Ke Ji

Fundus examination, conducted through the analysis of retinal images, plays a pivotal role in aiding the diagnosis of ophthalmic diseases. Currently, deep learning models have been widely applied to retinal image analysis, specifically retinal image segmentation, including vessels, optic cups, lesions, etc. , yielding promising outcomes. Nonetheless, the performance of these deep learning models experiences substantial degradation due to the domain shift between the distribution of the training images and unseen test images. In this paper, we focus on the challenging single-domain generalization (SDG), which aims to learn a generalized model on only one source domain, with the expectation that it performs well on unseen test domains. Our work is motivated by the observation that the main differences in retinal images from different domains primarily reside in their color variations, rather than changes in the shapes of objects. To this end, we present random color transformation (RCT) for SDG. RCT performs random linear transformations to each color channel of the training image. Through this approach, RCT can generate training images with rich and vibrant color representations, while preserving the structural information of objects in the images. Experiments are conducted over optic cup segmentation, retinal vessel segmentation, and diabetic retinopathy multi-lesion segmentation tasks, involving eight publicly available datasets. Experimental results show that the proposed RCT outperforms comparison SDG methods, achieving improvements of 5. 9%, 1. 3% and 3. 6% compared to the second-best method on the optic cup, vessel, and lesion segmentation tasks, respectively. The source code will be available at https: //github. com/guomugong/RCT.

Details DOI

IJCAI Conference 2024 Conference Paper

Recall, Retrieve and Reason: Towards Better In-Context Relation Extraction

Guozheng Li
Peng Wang
Wenjun Ke
Yikai Guo
Ke Ji
Ziyu Shang
Jiajun Liu
Zijie Xu

Relation extraction (RE) aims to identify relations between entities mentioned in texts. Although large language models (LLMs) have demonstrated impressive in-context learning (ICL) abilities in various tasks, they still suffer from poor performances compared to most supervised fine-tuned RE methods. Utilizing ICL for RE with LLMs encounters two challenges: (1) retrieving good demonstrations from training examples, and (2) enabling LLMs exhibit strong ICL abilities in RE. On the one hand, retrieving good demonstrations is a non-trivial process in RE, which easily results in low relevance regarding entities and relations. On the other hand, ICL with an LLM achieves poor performance in RE while RE is different from language modeling in nature or the LLM is not large enough. In this work, we propose a novel recall-retrieve-reason RE framework that synergizes LLMs with retrieval corpora (training examples) to enable relevant retrieving and reliable in-context reasoning. Specifically, we distill the consistently ontological knowledge from training datasets to let LLMs generate relevant entity pairs grounded by retrieval corpora as valid queries. These entity pairs are then used to retrieve relevant training examples from the retrieval corpora as demonstrations for LLMs to conduct better ICL via instruction tuning. Extensive experiments on different LLMs and RE datasets demonstrate that our method generates relevant and valid entity pairs and boosts ICL abilities of LLMs, achieving competitive or new state-of-the-art performance on sentence-level RE compared to previous supervised fine-tuning methods and ICL-based methods.

PDF Details DOI

AAAI Conference 2024 Conference Paper

Towards Continual Knowledge Graph Embedding via Incremental Distillation

Jiajun Liu
Wenjun Ke
Peng Wang
Ziyu Shang
Jinhua Gao
Guozheng Li
Ke Ji
Yanhe Liu

Traditional knowledge graph embedding (KGE) methods typically require preserving the entire knowledge graph (KG) with significant training costs when new knowledge emerges. To address this issue, the continual knowledge graph embedding (CKGE) task has been proposed to train the KGE model by learning emerging knowledge efficiently while simultaneously preserving decent old knowledge. However, the explicit graph structure in KGs, which is critical for the above goal, has been heavily ignored by existing CKGE methods. On the one hand, existing methods usually learn new triples in a random order, destroying the inner structure of new KGs. On the other hand, old triples are preserved with equal priority, failing to alleviate catastrophic forgetting effectively. In this paper, we propose a competitive method for CKGE based on incremental distillation (IncDE), which considers the full use of the explicit graph structure in KGs. First, to optimize the learning order, we introduce a hierarchical strategy, ranking new triples for layer-by-layer learning. By employing the inter- and intra-hierarchical orders together, new triples are grouped into layers based on the graph structure features. Secondly, to preserve the old knowledge effectively, we devise a novel incremental distillation mechanism, which facilitates the seamless transfer of entity representations from the previous layer to the next one, promoting old knowledge preservation. Finally, we adopt a two-stage training paradigm to avoid the over-corruption of old knowledge influenced by under-trained new knowledge. Experimental results demonstrate the superiority of IncDE over state-of-the-art baselines. Notably, the incremental distillation mechanism contributes to improvements of 0.2%-6.5% in the mean reciprocal rank (MRR) score. More exploratory experiments validate the effectiveness of IncDE in proficiently learning new knowledge while preserving old knowledge across all time steps.

PDF Details DOI

AAAI Conference 2023 Conference Paper

fmLRE: A Low-Resource Relation Extraction Model Based on Feature Mapping Similarity Calculation

Peng Wang
Tong Shao
Ke Ji
Guozheng Li
Wenjun Ke

Low-resource relation extraction (LRE) aims to extract relations from limited labeled corpora. Existing work takes advantages of self-training or distant supervision to expand the limited labeled data in the data-driven approaches, while the selection bias of pseudo labels may cause the error accumulation in subsequent relation classification. To address this issue, this paper proposes fmLRE, an iterative feedback method based on feature mapping similarity calculation to improve the accuracy of pseudo labels. First, it calculates the similarities between pseudo-label and real-label data of the same category in a feature mapping space based on semantic features of labeled dataset after feature projection. Then, it fine-tunes initial model according to the iterative process of reinforcement learning. Finally, the similarity is used as a threshold for screening high-precision pseudo-labels and the basis for setting different rewards, which also acts as a penalty term for the loss function of relation classifier. Experimental results demonstrate that fmLRE achieves the state-of-the-art performance compared with strong baselines on two public datasets.

PDF Details DOI

EAAI Journal 2021 Journal Article

Attention-based learning of self-media data for marketing intention detection

Zhihao Hou
Kun Ma
Yufeng Wang
Jia Yu
Ke Ji
Zhenxiang Chen
Ajith Abraham

In the context of natural language processing, accuracy of intention detection is the basis for subsequent research on human-machine speech interaction. However, the problem of ambiguity in word vectors reduces the accuracy of intent detection. Meantime, there is a disconnection between local features and global features as well, resulting in text feature extraction that cannot fully reflect semantic information. These issues are all barriers of intention detection. Therefore, this paper proposes an attention-based convolutional neural network for self-media data learning (called A-CNN) for marketing intention. We cascade the traditional CNN with the self-attention model in the Attention networks to form a new network structure called A-CNN, and put forward a fast feature extraction method based on skip-gram-based learning called FSLText, to represent the high-dimension word vectors in the A-CNN. On the premise of maintaining the advantages of the CNN, A-CNN can not only solve the problem of local and global features disconnection caused by the CNN pooling layer, but also avoid the increase of algorithm complexity. The Self-Attention mechanism in the Attention model can effectively optimize the weight of local features of the information in global features, and retain local features that are more useful for intention detection. A fast feature extraction method which is based on Skip-gram can retain the semantic and word order information of the text. The method is beneficial to the marketing intention detection. According to the experiment, our A-CNN, compared with traditional machine learning methods, can improve 12. 32% accuracy. Contrast to the dual-channel CNN, the accuracy rate is improved by 9. 68%, and compared with the ATT-CNN, it is improved by 9. 97%. On the F1 score, the A-CNN can improve the F1 score by about 9. 37% in comparison with the traditional machine learning methods, the accuracy rate is increased by 9. 68% compared with the dual-channel CNN, and 9. 68% in contrast with ATT-CNN. It illustrates that our A-CNN can effectively address semantic and feature selection for marketing intention detection.

Details DOI

EAAI Journal 2020 Journal Article

A CLSTM-TMN for marketing intention detection

Yufeng Wang
Kun Ma
Laura Garcia-Hernandez
Jing Chen
Zhihao Hou
Ke Ji
Zhenxiang Chen
Ajith Abraham

In recent years, neural network-based models such as machine learning and deep learning have achieved excellent results in text classification. On the research of marketing intention detection, classification measures are adopted to identify news with marketing intent. However, most of current news appears in the form of dialogs. There are some challenges to find potential relevance between news sentences to determine the latent semantics. In order to address this issue, this paper has proposed a CLSTM-based topic memory network (called CLSTM-TMN for short) for marketing intention detection. A ReLU-Neuro Topic Model (RNTM) is proposed. A hidden layer is constructed to efficiently capture the subject document representation, Potential variables are applied to enhance the granularity of subject model learning. We have changed the structure of current Neural Topic Model (NTM) to add CLSTM classifier. This method is a new combination ensemble both long and short term memory (LSTM) and convolution neural network (CNN). The CLSTM structure has the ability to find relationships from a sequence of text input, and the ability to extract local and dense features through convolution operations. The effectiveness of the method for marketing intention detection is illustrated in the experiments. Our detection model has a more significant improvement in F1 (7%) than other compared models.

Details DOI