检索结果-南通市图书馆

A language-independent neural network for event detection

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2018年第9期61卷 81-92页

作者： Xiaocheng FENG bing qin ting liu Computer Science and Technology Harbin Institute of Technology

Event detection remains a challenge because of the difficulty of encoding the word semantics in various contexts. Previous approaches have heavily depended on language-specific knowledge and preexisting natural language processing tools. However, not all languages have such resources and tools available compared with English language. A more promising approach is to automatically learn effective features from data, without relying on language-specific resources. In this study, we develop a language-independent neural network to capture both sequence and chunk information from specific contexts and use them to train an event detector for multiple languages without any manually encoded features. Experiments show that our approach can achieve robust, efficient and accurate results for various languages. In the ACE 2005 English event detection task, our approach achieved a 73.4% F-score with an average of 3.0% absolute improvement compared with state-of-the-art. Additionally, our experimental results are competitive for Chinese and Spanish.

关键词： nature language processing event detection neural networks representation learning

Encoding syntactic representations with a neural network for sentiment collocation extraction

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2017年第11期60卷 7-18页

作者： Yanyan ZHAO bing qin ting liu Department of Media Technology and Art Harbin Institute of Technology Department of Computer Science and Technology Harbin Institute of Technology

Sentiment collocation refers to the collocation of a target word and a polarity word. Sentiment collocation extraction aims to extract the targets and their modifying polarity words by analyzing the relationships between them. This can be regarded as a basic sentiment analysis task and is relevant in many practical applications. Previous studies relied mainly on the syntactic path, which is used to connect the target word and the polarity word. To deeply exploit the semantic information of the syntactic path, we propose two types of syntactic representation, namely, relation embedding and subtree embedding, to capture the latent semantic features. Relation embedding is used to represent the latent semantics between targets and their corresponding polarity words, and subtree embedding is used to explore the rich syntactic information for each word on the path. To combine the two types of syntactic representations, a neural network is constructed. We use a recursive neural network(RNN) to model the subtree embeddings, and then the subtree embedding and the word embedding are combined as the enhanced word representation for each word in the syntactic path. Finally,a convolutional neural network(CNN) is adopted to integrate the two types of syntactic representations to extract the sentiment collocations from reviews. Our experiments were conducted on six types of reviews, which included product domains(such as cameras and phones) and service domains(such as hotels and restaurants).The experimental results show that our proposed method can accurately capture the latent semantic features hidden behind the syntactic paths that neither the common feature-based methods nor the syntactic-path-based method can handle, and, further, that it significantly outperforms numerous baselines and previous methods.

关键词： sentiment collocation extraction sentiment analysis syntactic representation neural network re cursive neural network(RNN) convolutional neural network(CNN)

A survey of discourse parsing

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Frontiers of Computer Science 2022年第5期16卷 85-96页

作者： Jiaqi LI Ming liu bing qin ting liu School of Computer Science and Technology Harbin Institute of TechnologyHarbin 150001China Peng Cheng Laboratory Shenzhen 518052China

Discourse parsing is an important research area in natural language processing(NLP),which aims to parse the discourse structure of coherent *** this survey,we introduce several different kinds of discourse parsing tasks,mainly including RST-style discourse parsing,PDTB-style discourse parsing,and discourse parsing for multiparty *** these tasks,we introduce the classical and recent existing methods,especially neural network *** that,we describe the applications of discourse parsing for other NLP tasks,such as machine reading comprehension and sentiment ***,we discuss the future trends of the task.

关键词： discourse parsing discourse structure RST PDTB STAC

Topic hierarchy construction from heterogeneous evidence

在线全文

学校读者我要写书评

暂无评论

Frontiers of Computer Science 2016年第1期10卷 136-146页

作者： Han XUE bing qin ting liu Shen liu School of Computer Science and Technology Harbin Institute of Technology Harbin 150001 China Harbin Engineering University Library Harbin Engineering University Harbin 150001 China

Existing studies on hierarchy construction mainly focus on text corpora and indiscriminately mix numerous top- ics, thus increasing the possibility of knowledge acquisition bottlenecks and misconceptions. To address these problems and provide a comprehensive and in-depth representation of domain specific topics, we propose a novel topic hierar- chy construction method with real-time update. This method combines heterogeneous evidence from multiple sources in- cluding folksonomy and encyclopedia, separately in both ini- tial topic hierarchy construction and topic hierarchy improve- ment. Results of comprehensive experiments indicate that the proposed method significantly outperforms state-of-the- art methods （t-test, p-value 〈 0.000 1）; recall has particularly improved by 20.4% to 38.7%.

关键词： hierarchy construction Chinese topic hierarchy,folksonomy heterogeneous evidence hierarchy update

Truth Discovery with Memory Network

在线全文

学校读者我要写书评

暂无评论

Tsinghua Science and Technology 2017年第6期22卷 609-618页

作者： Luyang Li bing qin Wenjing Ren ting liu Research Center for Social Computing and Information Retrieval Harbin Institute of Technology Harbin 150001 China

Truth discovery aims to resolve conflicts among multiple sources and find the truth. Conventional methods for truth discovery mainly investigate the mutual effect between the reliability of sources and the credibility of statements. These methods use real numbers, which have a lower representation capability than vectors to represent the reliability. In addition, neural networks have not been used for truth discovery. In this work, we propose memory-network-based models to address truth discovery. Our proposed models use feedforward and feedback memory networks to learn the representation of the credibility of statements. Specifically, our models adopt a memory mechanism to learn the reliability of sources for truth prediction. The proposed models use categorical and continuous data during model learning by automatically assigning different weights to the loss function on the basis of their own effects. Experimental results show that our proposed models outperform state-of-the-art methods for truth discovery.

关键词： truth discovery memory networks source reliability

Text Difficulty Study:Do Machines Behave the Same as Humans Regarding Text Difficulty?

在线全文

学校读者我要写书评

暂无评论

Machine Intelligence Research 2024年第2期21卷 283-293页

作者： Bowen Chen Xiao Ding Yi Zhao Bo fu tingmao Lin bing qin ting liu Research Center for Social Computing and Information Retrieval Harbin Institute of TechnologyHarbin 150001China fundamental Technology Center China Construction Bank Financial Technology Co.Ltd.Beijing 100033China

With the emergence of pre-trained models,current neural networks are able to give task performance that is comparable to ***,we know little about the fundamental working mechanism of pre-trained models in which we do not know how they approach such performance and how the task is solved by the *** example,given a task,human learns from easy to hard,whereas the model learns ***,difficulty-insensitive learning leads to great success in natural language processing(NLP),but little attention has been paid to the effect of text difficulty in *** propose a human learning matching index(HLM Index)to investigate the effect of text *** results show:1)LSTM gives more human-like learning behavior than ***,UID-SuperLinear gives the best evaluation of text difficulty among four text difficulty *** nine tasks,some tasks’performance is related to text difficulty,whereas others are not.2)Model trained on easy data performs best in both easy and medium test data,whereas trained on hard data only performs well on hard test data.3)Train the model from easy to hard,leading to quicker convergence.

关键词： Cognition inspired natural language processing psycholinguistics explainability text difficulty curriculum learning

Quantifying the effects of long-term news on stock markets on the basis of the multikernel Hawkes process

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2021年第9期64卷 44-56页

作者： Xiao DING Jihao SHI Junwen DUAN bing qin ting liu Research Center for Social Computing and Information Retrieval

Recent studies have revealed that long-term financial news can affect stock markets. However,previous research mainly focuses on modeling the short-term effects of financial news and suffers from the weak ability of quantifying the time-decaying influence of financial news. To fill this gap, this study introduces the Hawkes process to estimate the time-decaying influence of long-term financial news. However, the performance of the conventional Hawkes process is sensitive to the choice of kernel functions. Hence, we propose a novel multikernel-powered Hawkes process framework, which uses multiple kernels to model different time-decaying rates, thus alleviating the instability of our proposed Hawkes process based prediction *** results show that the proposed framework yields state-of-the-art stock market prediction accuracies on 515 listed companies and gains more profits in market trading simulation compared with baseline methods. News-based stock prediction can complement studies on price-volume-based stock prediction.

关键词： Hawkes process multikernel function stock market prediction long-term news effects

A syntactic path-based hybrid neural network for negation scope detection

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Frontiers of Computer Science 2020年第1期14卷 84-94页

作者： Lydia LAZIB bing qin Yanyan ZHAO Weinan ZHANG ting liu Research Center for Social Computing and Information Retrieval Harbin Institute of TechnologyHarbin 150001China

The automatic detection of negation is a crucial task in a wide-range of natural language processing(NLP)applications,including medical data mining,relation extraction,question answering,and sentiment *** this paper,we present a syntactic path-based hybrid neural network architecture,a novel approach to identify the scope of negation in a *** hybrid architecture has the particularity to capture salient information to determine whether a token is in the scope or not,without relying on any human *** approach combines a bidirectional long shortterm memory(Bi-LSTM)network and a convolutional neural network(CNN).The CNN model captures relevant syntactic features between the token and the cue within the shortest syntactic path in both constituency and dependency parse *** Bi-LSTM learns the context representation along the sentence in both forward and backward *** evaluate our model on the Bioscope corpus,and get 90.82%F-score(78.31%PCS)on the abstract sub-corpus,outperforming features-dependent approaches.

关键词： natural language processing negation scope detection convolutional neural network recurrent neural network syntactic path

Recent advances in deep learning based sentiment analysis

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Science China(Technological Sciences) 2020年第10期63卷 1947-1970页

作者： YUAN JianHua WU Yang LU Xin ZHAO YanYan qin bing liu ting Research Center for Social Computing and Information Retrieval Harbin Institute of TechnologyHarbin 150001China Pengcheng Lab Shenzhen 518066China

Sentiment analysis is one of the most popular research areas in natural language *** is extremely useful in many applications,such as social media monitoring and *** application of deep learning based methods has dramatically changed the research strategies and improved the performance of many traditional sentiment analysis tasks,such as sentiment classification and aspect based sentiment ***,it also pushed the boundary of various sentiment analysis task,including sentiment classification of different text granularities and in different application scenarios,implicit sentiment analysis,multimodal sentiment analysis and generation of sentiment-bearing *** this paper,we give a brief introduction to the recent advance of the deep learning-based methods in these sentiment analysis tasks,including summarizing the approaches and analyzing the *** survey can be well suited for the researchers studying in this field as well as the researchers entering the field.

关键词： coarse-grained fine-grained implicit multi-modal generation