检索结果-内蒙古大学图书馆

An improved Bert learning model for E-commerce text entity extraction

JOURNAL OF SUPERCOMPUTING 2025年第4期81卷 1-28页

作者： Fan, Huiqiong Wan, Changxuan Jiangxi Univ Finance & Econ Sch Informat Management Nanchang 330032 Jiangxi Peoples R China Jiangxi Univ Finance & Econ Jiangxi Key Lab Data & Knowledge Engn Nanchang 330013 Jiangxi Peoples R China

In response to the intricate and non-standardized nature of e-commerce product descriptions, we propose an enhanced BERT-BiLSTM-CRF entity extraction model to address the limitations of existing models in accurately extracting necessary entities. BERT serves as the encoder, BiLSTM constructs the feature extractor, and CRF refines the model for improved accuracy. The e-commerce product text data are preprocessed and annotated according to real-world task requirements, and the model's performance is validated on both self-generated and MSRA public datasets. Experimental results show that the BERT-BiLSTM-CRF model achieves an F1 score of 98% and precision of 98% on the self-generated dataset, effectively catering to real-world e-commerce product information mining tasks. Furthermore, it achieves a relative balance between precision and recall, enabling more precise extraction of essential entities. The model's performance improvements are attributed to the integration of BERT's contextual understanding, BiLSTM's ability to capture long-range dependencies, and CRF's label sequence optimization, which collectively enhance the model's generalization and accuracy across diverse datasets. Compared to traditional models, the proposed BERT-BiLSTM-CRF model shows a 6.5% improvement in F1 score on the MSRA dataset and a 4% improvement in precision on the self-generated dataset, highlighting its superior capability in handling complex and non-standardized e-commerce text data.

关键词： Entity extraction E-commerce text big data Deep learning Bert learning model

来源：评论

学校读者我要写书评

暂无评论

A Multifocal Graph-Based Neural Network Scheme for Topic Event Extraction

引用

ACM TRANSACTIONS ON INFORMATION SYSTEMS 2025年第1期43卷 1-36页

作者： Wan, Qizhi Wan, Changxuan Xiao, Keli Hu, Rong Liu, Dexi Liao, Guoqiong Liu, Xiping Shuai, Yuxin Jiangxi Univ Finance & Econ Sch Comp & Artificial Intelligence Nanchang Peoples R China Jiangxi Key Lab Data & Knowledge Engn Nanchang Peoples R China SUNY Stony Brook Coll Business Stony Brook NY 11794 USA

Event extraction is a long-standing and challenging task in natural language processing, and existing studies mainly focus on extracting events within sentences. However, a significant problem that has not been carefully investigated is whether an "event topic" can be identified to represent the main aspects of extracted events. This article formulates the "topic event" extraction problem, aiming to identify a representative event from extracted ones. Specifically, after defining the topic event, we develop a multifocal graph-based framework to handle the extraction task. To enrich the associations of events and their tokens, we construct four event graphs, including the event subgraph and three event-associated graphs (i.e., event dependency parsing graph, event organization graph, and event share token graph), that reflect the internal and external structures of events, respectively. Subsequently, we design a multi-attention event-graph neural network to capture these event graph structures and improve event subgraph embedding. Finally, the output embeddings in the last layer of each channel are concatenated and fed into a fully connected network for topic event recognition. Extensive experiments validate the effectiveness of our method, and the results confirm its superiority over state-of-the-art baselines. In-depth analyses explore the essential factors (e.g., graph structures, attentions, feature generation method, etc.) determining the extraction performance.

关键词： Event Topic topic event extraction event graphs subgraph graph neural network

来源：评论

学校读者我要写书评

暂无评论

A Part-of-Speech Tagging Model Employing Word Clustering and Syntactic Parsing

引用

Chinese Journal of Electronics 2025年第1期23卷 109-114页

作者： Lichi Yuan School of Information Technology Jiangxi University of Finance and Economics Nanchang China Jiangxi Key Laboratory of Data and Knowledge Engineering Jiangxi University of Finance and Economics Nanchang China

Part-Of-Speech tagging is a basic task in the field of natural language processing. This paper builds a POS tagger based on improved Hidden Markov model, by employing word clustering and syntactic parsing model. Firstly, In order to overcome the defects of the classical HMM, Markov family model (MFM), a new statistical model was introduced. Secondly, to solve the problem of data sparseness, we propose a bottom-to-up hierarchical word clustering algorithm. Then we combine syntactic parsing with part-of-speech tagging. The Part-of-Speech tagging experiments show that the improved Part-Of-Speech tagging model has higher performance than Hidden Markov models (HMMs) under the same testing conditions, the precision is enhanced from 94.642% to 97.235%.

关键词：

来源：评论

学校读者我要写书评

暂无评论

LoRS: Efficient Low-Rank Adaptation for Sparse Large Language Model

arXiv

引用

arXiv 2025年

作者： Hu, Yuxuan Zhang, Jing Chen, Xiaodong Zhao, Zhe Li, Cuiping Chen, Hong School of Information Renmin University of China Beijing China Key Laboratory of Data Engineering and Knowledge Engineering Beijing China Engineering Research Center of Database and Business Intelligence Beijing China Tencent AI Lab Beijing China

Existing low-rank adaptation (LoRA) methods face challenges on sparse large language models (LLMs) due to the inability to maintain sparsity. Recent works introduced methods that maintain sparsity by augmenting LoRA techniques with additional masking mechanisms. Despite these successes, such approaches suffer from an increased memory and computation overhead, which affects efficiency of LoRA methods. In response to this limitation, we introduce LoRS, an innovative method designed to achieve both memory and computation efficiency when fine-tuning sparse LLMs. To mitigate the substantial memory and computation demands associated with preserving sparsity, our approach incorporates strategies of weight recompute and computational graph rearrangement. In addition, we also improve the effectiveness of LoRS through better adapter initialization. These innovations lead to a notable reduction in memory and computation consumption during the fine-tuning phase, all while achieving performance levels that outperform existing LoRA approaches. © 2025, CC BY.

关键词： Problem oriented languages

来源：评论

学校读者我要写书评

暂无评论

Improved head-driven statistical models for natural language parsing

引用

Journal of Central South University 2013年第10期20卷 2747-2752页

作者：袁里驰 School of Information Technology Jiangxi University of Finance and Economics Jiangxi Key Laboratory of Data and Knowledge Engineering Jiangxi University of Finance and Economics

Head-driven statistical models for natural language parsing are the most representative lexicalized syntactic parsing models, but they only utilize semantic dependency between words, and do not incorporate other semantic information such as semantic collocation and semantic category. Some improvements on this distinctive parser are presented. Firstly, "valency" is an essential semantic feature of words. Once the valency of word is determined, the collocation of the word is clear, and the sentence structure can be directly derived. Thus, a syntactic parsing model combining valence structure with semantic dependency is purposed on the base of head-driven statistical syntactic parsing models. Secondly, semantic role labeling(SRL) is very necessary for deep natural language processing. An integrated parsing approach is proposed to integrate semantic parsing into the syntactic parsing process. Experiments are conducted for the refined statistical parser. The results show that 87.12% precision and 85.04% recall are obtained, and F measure is improved by 5.68% compared with the head-driven parsing model introduced by Collins.

关键词： valence structure semantic dependency head-driven statistical syntactic parsing semantic role labeling

来源：评论

学校读者我要写书评

暂无评论

A Part-of-speech Tagging Model Employing Word Clustering and Syntactic Parsing

引用

Chinese Journal of Electronics 2014年第1期23卷 109-114页

作者： YUAN Lichi School of Information Technology Jiangxi University of Finance and Economics Jiangxi Key Laboratory of Data and Knowledge Engineering Jiangxi University of Finance and Economics

Part-Of-Speech tagging is a basic task in the field of natural language processing. This paper builds a POS tagger based on improved Hidden Markov model,by employing word clustering and syntactic parsing ***, In order to overcome the defects of the classical HMM, Markov family model(MFM), a new statistical model was introduced. Secondly, to solve the problem of data sparseness, we propose a bottom-to-up hierarchical word clustering algorithm. Then we combine syntactic parsing with part-of-speech tagging. The Part-ofSpeech tagging experiments show that the improved PartOf-Speech tagging model has higher performance than Hidden Markov models(HMMs) under the same testing conditions, the precision is enhanced from 94.642% to97.235%.

关键词： Part-of-speech tagging Hidden Markov model Markov family model Word clustering

来源：评论

学校读者我要写书评

暂无评论

Structural Join and Staircase Join Algorithms of Sibling Relationship

引用

Journal of Computer Science & Technology 2007年第2期22卷 171-181页

作者：万常选刘喜平 School of Information Technology Jiangxi University of Finance and Economics Nanchang 330013 China Jiangxi Key Laboratory of Data and Knowledge Engineering Nanchang 330013 China

The processing of XML queries can result in evaluation of various structural relationships. Efficient algorithms for evaluating ancestor-descendant and parent-child relationships have been proposed. Whereas the problems of evaluating preceding-sibling-following-sibling and preceding-following relationships are still open. In this paper, we studied the structural join and staircase join for sibling relationship. First, the idea of how to filter out and minimize unnecessary reads of elements using parent＇s structural information is introduced, which can be used to accelerate structural joins of parent-child and preceding-sibling-following-sibling relationships. Second, two efficient structural join algorithms of sibling relationship are proposed. These algorithms lead to optimal join performance： nodes that do not participate in the join can be judged beforehand and then skipped using B^＋-tree index. Besides, each element list joined is scanned sequentially once at most. Furthermore, output of join results is sorted in document order. We also discussed the staircase join algorithm for sibling axes. Studies show that, staircase join for sibling axes is close to the structural join for sibling axes and shares the same characteristic of high efficiency. Our experimental results not only demonstrate the effectiveness of our optimizing techniques for sibling axes, but also validate the efficiency of our algorithms. As far as we know, this is the first work addressing this problem specially.

关键词： XML query processing sibling relationship structural join staircase join

来源：评论

学校读者我要写书评

暂无评论

Preventing Recommendation Attack in Trust-Based Recommender Systems

引用

Journal of Computer Science & Technology 2011年第5期26卷 823-828页

作者：张富国 School of Information and Technology Jiangxi University of Finance and Economics Institute of Information Resource Management Jiangxi University of Finance and Economics Jiangxi Key Laboratory of Data and Knowledge Engineering Jiangxi University of Finance and Economics

Despite its success,similarity-based collaborative filtering suffers from some limitations,such as scalability,sparsity and recommendation *** work has shown incorporating trust mechanism into traditional collaborative filtering recommender systems can improve these *** argue that trust-based recommender systems are facing novel recommendation attack which is different from the profile injection attacks in traditional recommender *** the best of our knowledge,there has not any prior study on recommendation attack in a trust-based recommender *** analyze the attack problem,and find that ＂victim＂ nodes play a significant role in the ***,we propose a data provenance method to trace malicious users and identify the ＂victim＂ nodes as distrust users of recommender *** study of the defend method is done with the dataset crawled from Epinions website.

关键词： data lineage ＂victim＂ node attack trust propagation

来源：评论

学校读者我要写书评

暂无评论

Novel Semantics of the Top-k Queries on Uncertainly Fused Multi-Sensory data

引用

JOURNAL OF INFORMATION SCIENCE AND engineering 2015年第1期31卷 179-205页

作者： Liu, Dexi Jiangxi Univ Finance & Econ Sch Informat Technol Jiangxi Key Lab Data & Knowledge Engn Nanchang 330013 Peoples R China

Multiple sensors and sensor fusion are commonly used to get more accurate information. The intuitive method to store multi-sensory data is using uncertain database because the sensors are not precise enough. Hence, like the top-k queries in traditional database, the top-k queries in uncertain databases are quite popular and useful due to its wide application. Although there are lots of top-k query semantics, most of them return tuples, which does not make sense in some cases. We define two novel kinds of top-k query semantics in uncertain database, Uncertain x-kRanks queries (U-x-kRanks) and Global x-Top-k queries (G-x-Top-k), which return k x-tuples according to the score and the confidence of alternatives in x-tuples, respectively. Moreover, in order to reduce the search space, we present an efficient algorithm to process U-x-kRanks queries and G-x-Top-k queries. Comprehensive experiments and analysis on different artificial data sets demonstrate the effectiveness of the proposed strategies.

关键词： multi-sensory data top-k query semantics x-tuple Uncertain x-kRanks queries Global x-Top-k queries

来源：评论

学校读者我要写书评

暂无评论

2-Tuple linguistic hybrid arithmetic aggregation operators and application to multi-attribute group decision making

引用

knowledge-BASED SYSTEMS 2013年 45卷 31-40页

作者： Wan, Shu-Ping Jiangxi Univ Finance & Econ Coll Informat Technol Nanchang 330013 Jiangxi Peoples R China Jiangxi Univ Finance & Econ Jiangxi Key Lab Data & Knowledge Engn Nanchang 330013 Jiangxi Peoples R China

The focus of this paper is on multi-attribute group decision making (MAGDM) problems in which the attribute values, attribute weights, and expert weights are all in the form of 2-tuple linguistic information, which are solved by developing a new decision method based on 2-tuple linguistic hybrid arithmetic aggregation operator. First, the operation laws for 2-tuple linguistic information are defined and the related properties of the operation laws are studied. Hereby some hybrid arithmetic aggregation operators with 2-tuple linguistic information are developed, involving the 2-tuple hybrid weighted arithmetic average (THWA) operator, the 2-tuple hybrid linguistic weighted arithmetic average (T-HLWA) operator, and the extended 2-tuple hybrid linguistic weighted arithmetic average (ET-HLWA) operator. In the proposed decision method, the individual overall preference values of alternatives are derived by using the extended 2-tuple weighted arithmetic average operator (ET-WA). Utilized the ET-HLWA operator, all the individual overall preference values of alternatives are further integrated into the collective ones of alternatives, which are used to rank the alternatives. A real example of personnel selection is given to illustrate the developed method and the comparison analyses demonstrate the universality and flexibility of the method proposed in this paper. (C) 2013 Elsevier B.V. All rights reserved.

关键词： Multi-attribute group decision making Linguistic preference 2-Tuple linguistic information Hybrid aggregation operator Arithmetic average operators Geometric average operators

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：