Focus-sensitive relation disambiguation for implicit discourse relation detection

We study implicit discourse relation detection, which is one of the most challenging tasks in the field of discourse analysis. We specialize in ambiguous implicit discourse relation, which is an imperceptible linguistic phenomenon and therefore difficult to identify and eliminate. In this paper, we first create a novel task named implicit discourse relation disambiguation (IDRD). Second, we propose a focus-sensitive relation disambiguation model that affirms a truly-correct relation when it is triggered by focal sentence constituents. In addition, we specifically develop a topicdriven focus identification method and a relation search system (RSS) to support the relation disambiguation. Finally, we improve current relation detection systems by using the disambiguation model. Experiments on the penn discourse treebank (PDTB) show promising improvements.     

运用多层注意力神经网络识别中文隐式篇章关系

中文隐式篇章关系识别是一个具有挑战性的任务,其难点在于如何捕获论元的语义信息。该文提出了一个模拟人类双向阅读和重复阅读过程的三层注意力网络模型(TLAN)用于识别中文隐式篇章关系。首先,使用Self-Attention层对论元进行编码;然后,通过细粒度的Interactive Attention层模拟双向阅读过程以生成包含交互信息的论元表示,并且通过非线性变换获得论元对信息的外部记忆;最后,通过包含外部记忆的注意力层来模拟重复阅读过程,在论元对记忆的引导下生成论元的最终表示。在中文篇章树库(CDTB)上进行的隐式篇章关系识别实验结果显示,该文提出的模型TLAN在Micro-F1和Macro-F1上超过了多个基准模型。     

英语语篇结构分析研究综述

全面系统地分析了英语语篇结构分析的相关理论、语料资源及国内外的相关研究成果,给出了语篇结构分析的研究趋势,为英语和汉语语篇结构分析研究做了基础性的工作。     

基于门控记忆网络的汉语篇章主次关系识别方法

篇章分析是自然语言理解的基础。作为篇章分析的重要任务之一,汉语主次关系识别还处于探索阶段。该文提出了一种基于门控记忆网络(GMN)的汉语篇章主次关系识别方法。该方法首先使用Bi-LSTM和CNN分别获取每个篇章单元的全局信息和局部信息。然后,融合两部分篇章单元信息并从中计算得到一个门控单元。最后,使用这个门控单元捕获各个篇章单元相对于篇章整体来说相对重要的特征表示,从而识别出核心篇章单元。在Chinese Discourse Treebank(CDTB)语料库上的实验显示,和最好的基准系统相比,该文的方法在宏平均F1、微平均F1值上均得到了提高。     

Plan parsing for intended response recognition in discourse1

In a discourse the hearer must recognize the response intended by the speaker. To perform this recognition, the hearer must ascertain what plans the speaker is undertaking and how the utterances in the discourse further that plan. To do so, the hearer can parse the initial intentions (recoverable from the utterance) and recognize the plans the speaker has in mind and intends the hearer to know about. This paper reports on a theory of parsing the intentions in discourse. It also discusses the role of another aspect of discourse, discourse markers, that are valuable to intended response recognition.     

基于多层局部推理的汉语篇章关系及主次联合识别

篇章关系识别是篇章分析的核心组成部分。汉语中,缺少显式连接词的隐式篇章关系占比很高,篇章关系识别更具挑战性。该文给出了一个基于多层局部推理的汉语篇章关系及主次联合识别方法。该方法借助双向LSTM和多头自注意力机制进行篇章关系对应论元的表征;进一步借助软对齐方式获取论元间局部语义的推理权重,形成论元间交互语义信息的表征;再将两类信息结合进行篇章关系的局部推理,并通过堆叠多层局部推理部件构建了汉语篇章关系及主次联合识别框架,在CDTB语料库上的关系识别F₁值达到了67.0%。该文进一步将该联合识别模块嵌入一个基于转移的篇章解析器,在自动生成的篇章结构下进行篇章关系及主次的联合分析,形成了完整的汉语篇章解析器。     

中文篇章级句间语义关系识别

篇章句间关系识别(Discourse Relation Recognition)是篇章分析的重要内容,该文对中文篇章句间关系识别任务进行初步探索,包括显式篇章句间关系识别与隐式篇章句间关系识别两类任务。针对显式篇章句间关系,我们提出基于关联词规则的方法进行识别,取得了很好的效果;针对隐式篇章句间关系,我们抽取词汇、句法、语义等特征,采用有指导模型进行识别。该文的分析和实验结果为后续研究提供了参考和基本对照系统。     

基于篇章主题的中文宏观篇章主次关系识别方法

篇章分析是自然语言处理领域研究的热点和重点。作为篇章分析的任务之一,篇章主次关系研究篇章的主要和次要内容,从而更好地理解和把握篇章的核心内容。该文重点研究宏观领域的中文篇章主次关系,提出了一种基于篇章主题的中文宏观篇章主次关系识别方法。该方法利用篇章单元间、篇章单元与篇章主题间的语义交互来识别主次关系,并有选择地应用篇章主题信息,有效提高了主次关系核心的识别。在中文宏观汉语篇章树库(MCDTB)上的实验结果显示,该方法优于目前性能最好的基准系统。     

基于多元特征的分块人物关系识别系统

随着互联网的飞速发展,大量的文本信息被分享到网上,如何在海量的网络信息中提取出可靠性较高的人物关系已成为信息抽取领域中的一个重要研究课题。为深入进行人物关系识别任务在中文方面的研究,提出了基于多元特征的分块人物关系识别系统,设计了较为完备的特征池,包括词袋特征、相关频率特征、依存树(DT)特征、命名实体识别(NER)特征等,为不同的关系从特征池中选择效果最佳的特征集合,并实验了多种基于有监督的机器学习分类算法。本系统在2015年中国机器学习会议竞赛(CCML Competition)举办的两个任务(Task1是从单个新闻标题中判定给定人物的关系;Task2是从多个新闻标题中判定人物的关系)的数据集上分别取得了75.68%和76.58%的MacroF1值,均位列参赛成绩的第一名。     

基于时序和上下文特征的中文隐式情感分类模型

对社交网络上的海量文本信息进行情感分析可以更好地挖掘网民行为规律,从而帮助决策机构了解舆情倾向以及帮助商家改善服务质量。由于不存在关键情感特征、表达载体形式和文化习俗等因素的影响,中文隐式情感分类任务比其他语言更加困难。已有的中文隐式情感分类方法以卷积神经网络（CNN）为主,这些方法存在着无法获取词语的时序信息和在隐式情感判别中未合理利用上下文情感特征的缺陷。为了解决以上问题,采用门控卷积神经网络（GCNN）提取隐式情感句的局部重要信息,采用门控循环单元（GRU）网络增强特征的时序信息;而在隐式情感句的上下文特征处理上,采用双向门控循环单元（BiGRU）+注意力机制（Attention）的组合提取重要情感特征;在获得两种特征后,通过融合层将上下文重要特征融入到隐式情感判别中;最后得到的融合时序和上下文特征的中文隐式情感分类模型被命名为GGBA。在隐式情感分析评测数据集上进行实验,结果表明所提出的GGBA模型在宏平均准确率上比普通的文本CNN即TextCNN提高了3.72%、比GRU提高了2.57%、比中断循环神经网络（DRNN）提高了1.90%,由此可见, GGBA模型在隐式情感分析任务中比基础模型获得了更好的分类性能。     

Evolutionary joint selection to improve human action recognition with RGB-D devices

Interest in RGB-D devices is increasing due to their low price and the wide range of possible applications that come along. These devices provide a marker-less body pose estimation by means of skeletal data consisting of 3D positions of body joints. These can be further used for pose, gesture or action recognition. In this work, an evolutionary algorithm is used to determine the optimal subset of skeleton joints, taking into account the topological structure of the skeleton, in order to improve the final success rate. The proposed method has been validated using a state-of-the-art RGB action recognition approach, and applying it to the MSR-Action3D dataset. Results show that the proposed algorithm is able to significantly improve the initial recognition rate and to yield similar or better success rates than the state-of-the-art methods.     

空袭目标识别的证据理论方法

针对具有多个特征指标的空袭目标识别问题,利用向量夹角余弦提出了一种新的融合方法。该方法对参考模式和传感器获取信息的特征化向量进行规范化,利用传感器获取信息特征化向量与参考模式向量之间的夹角余弦,作为证据对各可能目标识别对象的基本概率指派,根据证据组合规则进行融合得到目标识别结果。实例分析验证了方法的有效性。     

基于证据可信度的综合目标识别方法

为了解决多传感器综合目标识别中不同等级信息源数据的融合问题,在研究D—S证据理论的基础上,引入证据可信度矩阵。依据可信度矩阵对证据进行转化,使之可以用传统的方法进行证据融合。将这种方法应用到等级不同的多传感器综合目标识别中,可以解决传统证据理论只能进行相同等级传感器目标识别的难题。仿真实验表明:该方法提高了目标识别的准确性和有效性。     

基于D-S证据理论的掌纹识别

研究了掌纹识别问题,对掌纹图像特征提取、多特征的融合技术作了一定程度的探讨。采用数学形态学方法提取掌纹线特征;基于Gabor滤波器描述掌纹图像的纹理特征。利用掌纹的线特征和纹理特征两个信息分别作两个分类器的特征,利用模糊规则求出各分类器的基本概率分配函数,最后利用D-S证据理论的合成法则对两个分类器的结果进行融合判决。实验结果表明,这种方法是有效的,可行的。     

基于加权D-S证据理论的分布式多传感器目标识别

针对分布式多传感器环境下的目标识别问题,提出了一种基于加权D-S证据理论组合规则的决策融合方法。分析了多传感器目标识别系统的信息模型,指出传感器的决策可信度由其被支持度及与目标间的距离确定。将该可信度体现为加权D-S证据理论组合规则中的证据权值,综合考虑传感器支持度及其与目标距离,给出了权值确定方法。仿真实验证明方法提高了融合效率,可较快完成识别任务。     

A unified approach to the recognition of complex actions from sequences of zone-crossings

We present a method for the recognition of complex actions. Our method combines automatic learning of simple actions and manual definition of complex actions in a single grammar. Contrary to the general trend in complex action recognition that consists in dividing recognition into two stages, our method performs recognition of simple and complex actions in a unified way. This is performed by encoding simple action HMMs within the stochastic grammar that models complex actions. This unified approach enables a more effective influence of the higher activity layers into the recognition of simple actions which leads to a substantial improvement in the classification of complex actions. We consider the recognition of complex actions based on person transits between areas in the scene. As input, our method receives crossings of tracks along a set of zones which are derived using unsupervised learning of the movement patterns of the objects in the scene. We evaluate our method on a large dataset showing normal, suspicious and threat behaviour on a parking lot. Experiments show an improvement of ~ 30% in the recognition of both high-level scenarios and their composing simple actions with respect to a two-stage approach. Experiments with synthetic noise simulating the most common tracking failures show that our method only experiences a limited decrease in performance when moderate amounts of noise are added.     

中文电子病历中的时间关系识别

中文电子病历中的时间关系包括句内时间关系和句间时间关系,其中,句内时间关系包括句内事件-事件的时间关系和句内事件-时间的时间关系,句间时间关系即是句间事件-事件的时间关系。把中文电子病历文本中的时间关系识别转化成实体对分类问题,针对句内时间关系的识别,制定了高准确率的启发式规则,并设计了基本特征、短语句法特征、依存特征和其他特征,训练分类器缓解句内时间关系的识别错误;针对句间时间关系的识别,在高准确率的启发式规则之外,设计了基本特征、短语句法特征和其他特征,训练分类器减少句间时间关系的识别错误。实验结果表明,当分别使用支持向量机（SVM）、SVM和随机森林（RF）算法时,所提方法在句内事件-事件、句内事件-时间和句间事件-事件的时间关系识别上的效果最好,其F₁值分别达到了84.0%、85.6%和63.5%。     

Conditional models for contextual human motion recognition

We describe algorithms for recognizing human motion in monocular video sequences, based on discriminative conditional random fields (CRFs) and maximum entropy Markov models (MEMMs). Existing approaches to this problem typically use generative structures like the hidden Markov model (HMM). Therefore, they have to make simplifying, often unrealistic assumptions on the conditional independence of observations given the motion class labels and cannot accommodate rich overlapping features of the observation or long-term contextual dependencies among observations at multiple timesteps. This makes them prone to myopic failures in recognizing many human motions, because even the transition between simple human activities naturally has temporal segments of ambiguity and overlap. The correct interpretation of these sequences requires more holistic, contextual decisions, where the estimate of an activity at a particular timestep could be constrained by longer windows of observations, prior and even posterior to that timestep. This would not be computationally feasible with a HMM which requires the enumeration of a number of observation sequences exponential in the size of the context window. In this work we follow a different philosophy: instead of restrictively modeling the complex image generation process – the observation, we work with models that can unrestrictedly take it as an input, hence condition on it. Conditional models like the proposed CRFs seamlessly represent contextual dependencies and have computationally attractive properties: they support efficient, exact recognition using dynamic programming, and their parameters can be learned using convex optimization. We introduce conditional graphical models as complementary tools for human motion recognition and present an extensive set of experiments that show not only how these can successfully classify diverse human activities like walking, jumping, running, picking or dancing, but also how they can discriminate among subtle motion styles like normal walks and wander walks.     

面向煤矿的实体识别与关系抽取模型

针对煤矿领域知识抽取中存在的术语嵌套、一词多义,抽取任务间存在误差传播等问题,提出了一种深层注意力模型框架。首先,使用标注策略联合学习两项知识抽取子任务,以解决误差传播的问题;其次,提出结合多种词向量信息的投影方法,以缓解煤矿领域术语抽取中的一词多义的问题;然后,设计深度特征提取网络,并提出深层注意力模型及两种模型增强方案来充分提取语义信息;最后,对模型的分类层进行研究,以在保证抽取效果的前提下最大限度地简化模型。实验结果表明,在煤矿领域语料上,相较于编码-解码结构的最好模型,所提模型的F1值有了1.5个百分点的提升,同时模型训练速度几乎提高至原来的3倍。该模型可有效地完成煤矿领域术语抽取以及术语关系抽取这两项知识抽取子任务。     

一种有效的基于证据理论的离线签名识别方法

提出了一种有效的基于证据理论的离线签名识别方法。从签名图像的3种信息载体中提取出4种特征,利用所提取的4种特征分别构造基于证据理论的k-NN分类器对签名图像进行初步识别,将各k-NN分类器的输出作为证据,用改进的证据理论合成公式融合不同分类器的输出得到最终识别结果。结果表明：该识别方法能有效地提高离线签名的识别率。