Automatic Event Trigger Word Extraction in Chinese Event

As a basic unit of knowledge representation and an important means for information organization, event has drawn growing number of people’s attention, the research of event identification and extraction in natural language processing field is an important research topic in information extraction area, the recognition and extraction of event trigger word plays a decisive role in event identification and extraction. In this paper, the authors make experiment in Chinese Event Corpus CEC, and present a method of extracting event trigger word automatically that combines extended trigger word table and machine learning. The experiment result shows that the F-score of extracting event trigger word. can reach 71.2% by using this method.

Improving the Collocation Extraction Method Using an Untagged Corpus for Persian Word Sense Disambiguation

Word sense disambiguation is used in many natural language processing fields. One of the ways of disambiguation is the use of decision list algorithm which is a supervised method. Supervised methods are considered as the most accurate machine learning algorithms but they are strongly influenced by knowledge acquisition bottleneck which means that their efficiency depends on the size of the tagged training set, in which their preparation is difficult, time-consuming and costly. The proposed method in this article improves the efficiency of this algorithm where there is a small tagged training set. This method uses a statistical method for collocation extraction from a big untagged corpus. Thus, the more important collocations which are the features used for creation of learning hypotheses will be identified. Weighting the features improves the efficiency and accuracy of a decision list algorithm which has been trained with a small training corpus.

Apriori and N-gram Based Chinese Text Feature Extraction Method

A feature extraction, which means extracting the representative words from a text, is an important issue in text mining field. This paper presented a new Apriori and N-gram based Chinese text feature extraction method, and analyzed its correctness and performance. Our method solves the question that the exist extraction methods cannot find the frequent words with arbitrary length in Chinese texts. The experimental results show this method is feasible.

Text Rank for Domain Specific Using Field Association Words

Text Rank is a popular tool for obtaining words or phrases that are important for many Natural Language Processing (NLP) tasks. This paper presents a practical approach for Text Rank domain specific using Field Association (FA) words. We present the keyphrase separation technique not for a single document, although for a particular domain. The former builds a specific domain field. The second collects a list of ideal FA terms and compounds FA terms from the specific domain that are considered to be contender keyword phrases. Therefore, we combine two-word node weights and field tree relationships into a new approach to generate keyphrases from a particular domain. Studies using the changed approach to extract key phrases demonstrate that the latest techniques including FA terms are stronger than the others that use normal words and its precise words reach 90%.

Content Feature Extraction-based Hybrid Recommendation for Mobile Application Services

The number of mobile application services is showing an explosive growth trend,which makes it difficult for users to determine which ones are of interest.Especially,the new mobile application services are emerge continuously,most of them have not be rated when they need to be recommended to users.This is the typical problem of cold start in the field of collaborative filtering recommendation.This problem may makes it difficult for users to locate and acquire the services that they actually want,and the accuracy and novelty of service recommendations are also difficult to satisfy users.To solve this problem,a hybrid recommendation method for mobile application services based on content feature extraction is proposed in this paper.First,the proposed method in this paper extracts service content features through Natural Language Processing technologies such as word segmentation,part-of-speech tagging,and dependency parsing.It improves the accuracy of describing service attributes and the rationality of the method of calculating service similarity.Then,a language representation model called Bidirectional Encoder Representation from Transformers(BERT)is used to vectorize the content feature text,and an improved weighted word mover’s distance algorithm based on Term Frequency-Inverse Document Frequency(TFIDF-WMD)is used to calculate the similarity of mobile application services.Finally,the recommendation process is completed by combining the item-based collaborative filtering recommendation algorithm.The experimental results show that by using the proposed hybrid recommendation method presented in this paper,the cold start problem is alleviated to a certain extent,and the accuracy of the recommendation result has been significantly improved.

Algorithm of the Real-Time Extraction Image for Vehicle

An algorithm applied to a real-time extraction image of vehicle is introduced. The algorithm include an image processing with a binarzation method, image grab for a vehicle with high speed, character isolator one by one, and neural network algorithm. The techniques include vehicles sensing, image garb control, vehicle license location, lighting and optic character recognition. The system is much more robust and faster than the traditional thresholding method.

基于深度学习与规则匹配的Word文档实体识别与属性抽取融合算法及其在油气勘探领域中的应用

针对油气勘探领域积累的大量历史Word文档中的知识未能很好地应用于实际生产和研究的问题,提出了一种基于深度学习与规则匹配的融合算法,从文档中抽取文本,从表格中抽取实体值与属性值。抽取文档中的内容,分析所要抽取的实体和属性的语义规则以及在文档、表格的位置规则,并以源数据的相关语料作为训练数据,基于BiLSTM+CRF模型完成实体、属性抽取模型的训练,结果准确率和召回率均在97%以上。将深度学习和规则匹配模型相结合,可实现实体和属性的高效准确抽取。目前,该融合算法在塔里木智能方案编写项目上取得了良好的应用效果,具有广阔的发展前景。

面向Word和PDF文档的图像提取软件

针对文档图像查重、以图搜索文档等应用对于文档图像提取的需求,文中归纳和总结了Word和PDF文档图像的提取方法。这两类文档图像的提取方法都是从文件结构出发,定位并提取出图像数据,然后将数据保存成指定格式的图像。在PyCharm环境下采用Python语言对提取方法进行封装,设计并实现了一个面向Word和PDF文档的图像提取软件,实验验证了文档图像提取方法的有效性。

基于MMR和WordNet的新闻文本摘要生成研究

针对新闻文本摘要提取过程中,传统抽取式算法存在对文本内容概括不全面、摘要内容冗余、关键词提取时未考虑异词同义等问题,提出了一种基于最大边界相关算法(MMR)和词汇语义网(WordNet)的新闻文本摘要生成算法--WMMR.该算法综合考虑文本相似度、关键词、句子位置信息、线索词等特征对句子权重的影响,从而优化MMR算法中的句子得分,并在计算关键词得分时引入WordNet合并同义词.在NLPCC2017公开数据集上验证本文算法的有效性,结果表明WMMR算法的ROUGE值相较于TextRank算法提升4个百分点,相较于MMR算法提升7个百分点.在神策杯2018与SogouCS公开数据集上验证本文算法的普适性,结果表明WMMR算法的ROUGE值相较于传统TextRank,MMR等算法均有提升,证明WMMR算法有效提升了生成摘要的质量.

Aspect-Based Sentiment Classification Using Deep Learning and Hybrid of Word Embedding and Contextual Position

Aspect-based sentiment analysis aims to detect and classify the sentiment polarities as negative,positive,or neutral while associating them with their identified aspects from the corresponding context.In this regard,prior methodologies widely utilize either word embedding or tree-based rep-resentations.Meanwhile,the separate use of those deep features such as word embedding and tree-based dependencies has become a significant cause of information loss.Generally,word embedding preserves the syntactic and semantic relations between a couple of terms lying in a sentence.Besides,the tree-based structure conserves the grammatical and logical dependencies of context.In addition,the sentence-oriented word position describes a critical factor that influences the contextual information of a targeted sentence.Therefore,knowledge of the position-oriented information of words in a sentence has been considered significant.In this study,we propose to use word embedding,tree-based representation,and contextual position information in combination to evaluate whether their combination will improve the result’s effectiveness or not.In the meantime,their joint utilization enhances the accurate identification and extraction of targeted aspect terms,which also influences their classification process.In this research paper,we propose a method named Attention Based Multi-Channel Convolutional Neural Net-work(Att-MC-CNN)that jointly utilizes these three deep features such as word embedding with tree-based structure and contextual position informa-tion.These three parameters deliver to Multi-Channel Convolutional Neural Network(MC-CNN)that identifies and extracts the potential terms and classifies their polarities.In addition,these terms have been further filtered with the attention mechanism,which determines the most significant words.The empirical analysis proves the proposed approach’s effectiveness compared to existing techniques when evaluated on standard datasets.The experimental results represent our approach outperforms in the F1 measure with an overall achievement of 94%in identifying aspects and 92%in the task of sentiment classification.

Word Sense Disambiguation Based Sentiment Classification Using Linear Kernel Learning Scheme

Word Sense Disambiguation has been a trending topic of research in Natural Language Processing and Machine Learning.Mining core features and performing the text classification still exist as a challenging task.Here the features of the context such as neighboring words like adjective provide the evidence for classification using machine learning approach.This paper presented the text document classification that has wide applications in information retrieval,which uses movie review datasets.Here the document indexing based on controlled vocabulary,adjective,word sense disambiguation,generating hierarchical cate-gorization of web pages,spam detection,topic labeling,web search,document summarization,etc.Here the kernel support vector machine learning algorithm helps to classify the text and feature extract is performed by cuckoo search opti-mization.Positive review and negative review of movie dataset is presented to get the better classification accuracy.Experimental results focused with context mining,feature analysis and classification.By comparing with the previous work,proposed work designed to achieve the efficient results.Overall design is per-formed with MATLAB 2020a tool.

基于Word2Vec及TextRank算法的长文档摘要自动生成研究

近年来,如何从大量信息中提取关键信息已成为一个急需解决的问题。针对中文专利长文档,提出一种结合Word2Vec和TextRank的专利生成算法。首先利用Python Jieba技术对中文专利文档进行分词,利用停用词典去除无意义的词;其次利用Word2Vec算法进行特征提取,并利用WordCloud对提取的关键词进行可视化展示;最后利用TextRank算法计算语句间的相似度,生成摘要候选句,根据候选句的权重生成该专利文档的摘要信息。实验表明,采用Word2Vec和TextRank生成的专利摘要质量高,概括性也强。

基于深度字词融合的小麦种质信息实体关系联合抽取

为获得结构化的小麦品种表型和遗传描述,针对非结构化小麦种质数据中存在的实体边界模糊以及关系重叠问题,提出一种基于深度字词融合的小麦种质信息实体关系联合抽取模型WGIE-DCWF(wheat germplasm information extraction model based on deep character and word fusion)。模型编码层通过深度字词融合和上下文语义特征融合,提高密集实体特征识别能力;模型三元组抽取层建立层叠指针网络,提高重叠关系的提取能力。在小麦种质数据集和公开数据集上的一系列对比实验结果表明,WGIE-DCWF模型能够有效提高小麦种质数据实体关系联合抽取效果,同时拥有较好的泛化性,可以为小麦种质信息知识库构建提供技术支撑。

基于组合相似度动态聚类和词熵的网络话题在线检测

[研究目的]为实现网络热点话题的在线检测,提升增量式聚类算法的聚类效果,提出了基于组合相似度的动态聚类算法,同时通过计算词熵实现主题词提取和演化跟踪。[研究方法]通过CIFG-BiLSTM-CRF模型实现文本的命名实体识别,计算文本与话题的实体相似度,再取文本词向量与话题中心余弦相似度的最大值作为词向量相似度,二者结合判断文本所属话题。在聚类过程中利用时间窗口策略实现话题中心和成员文本的动态更新。同时,计算文本词熵,生成话题的词熵和列表,实现话题主题词提取和演化跟踪。实验以新冠疫情新闻为数据实现话题在线检测,并展示了话题主题词的演化和跟踪过程。[研究结论]实验表明,与传统相似度计算方法相比,组合相似度能够获得更好的聚类效果,聚类过程中提取出的话题主题词也正确地反映了原始数据的热点话题内容。

基于BERT的两次注意力机制远程监督关系抽取

针对词向量语义信息不完整以及文本特征抽取时的一词多义问题,提出基于BERT(Bidirectional Encoder Representation from Transformer)的两次注意力加权算法(TARE)。首先,在词向量编码阶段,通过构建Q、K、V矩阵使用自注意力机制动态编码算法,为当前词的词向量捕获文本前后词语义信息;其次,在模型输出句子级特征向量后,利用定位信息符提取全连接层对应参数,构建关系注意力矩阵;最后,运用句子级注意力机制算法为每个句子级特征向量添加不同的注意力分数,提高句子级特征的抗噪能力。实验结果表明:在NYT-10m数据集上,与基于对比学习框架的CIL(Contrastive Instance Learning)算法相比,TARE的F1值提升了4.0个百分点,按置信度降序排列后前100、200和300条数据精准率Precision@N的平均值(P@M)提升了11.3个百分点;在NYT-10d数据集上,与基于注意力机制的PCNN-ATT(Piecewise Convolutional Neural Network algorithm based on ATTention mechanism)算法相比,精准率与召回率曲线下的面积(AUC)提升了4.8个百分点,P@M值提升了2.1个百分点。在主流的远程监督关系抽取(DSER)任务中,TARE有效地提升了模型对数据特征的学习能力。

基于词模式规则的轻量级日志模板提取方法

传统基于规则的日志解析方法针对每类日志需单独编写规则,且随着系统更新,出现新的日志模式时,需人工再次干预;基于深度学习的日志解析方法虽准确率高,但计算复杂度高。为解决日志解析方法人力成本和计算复杂度高的问题,文中提出一种基于词模式规则的轻量级日志模板提取方法,该方法由初始规则集生成、词模式规则应用、潜在错误样本发掘三个部分构成。首先,原始日志基于自适应随机抽样获取彼此间相似度较低的代表性日志;然后,基于专家反馈提取初始词模式规则集,在词模式规则应用模块对原始日志进行处理并提取日志模板;最后,在潜在错误样本发掘模块检查生成的日志模板聚类,发现潜在的错误分类样本并对其进行规则集更新。经过实验验证,在16个公开日志数据集上,文中方法的平均准确度达到97.8%,与基于深度学习的日志解析算法准确度基本持平;在计算效率方面,文中方法的单线程解析速度达到每秒20000条,且随着可用内核数量的增加,性能持续提升,满足系统日志的故障诊断和安全分析需求。

基于提示学习的篇章级事件论元抽取方法研究

事件论元抽取是指在自然语言文本中识别出事件论元及其对应的角色,是事件抽取的关键。传统事件论元抽取方法将抽取范围局限在单个句子中,在面对长文本中论元分散的情况时表现不佳。近年来,有研究者提出基于提示学习的篇章级事件论元抽取方法,能根据提示信息在输入文本中获取事件论元,实现事件论元抽取。然而现有基于提示学习的方法大多是由人工手动构建提示模板,模板结构固定容易导致论元抽取错误。针对以上不足,该文在以往基于提示学习研究的基础上,提出以文本触发词为关键实现模板自动构建的方法,并在输入文本中融入事件角色语义信息,使模型能更好地捕获文本语义特征,提高事件论元抽取准确率。在篇章级数据集RAMS上的实验结果表明,该模型在事件论元识别和事件论元分类的F1值分别达到54.3%和48.1%,相比最优的基准方法分别提升了1.8百分点和1.2百分点,验证了模型的有效性。

基于卷积神经网络的医疗护理实体关系抽取

针对医疗护理领域知识复杂性强、数据量大以及对准确度要求较高的问题,该研究提出一种基于卷积神经网络的医疗护理学实体关系抽取方法,实现对护理学语义关系的细粒度文本挖掘。该研究构建了医疗护理学语料标注系统,通过将医疗语料转化为向量特征矩阵,实现了对医疗语料的自动过滤和标注。通过向神经网络模型嵌入所构建的医疗关系语料库,一定程度上提高了模型疾病分类的准确度。在医疗护理学数据集上的实验表明,基于卷积神经网络的模型在指标精确度、召回率、F1值可达到89.78%、87.59%、89.77%。综上所述,该研究提出的基于卷积神经网络的医疗护理学实体关系抽取方法能够有效地抽取医疗语料数据中的实体关系,优于传统的实体关系抽取模型。

基于深度学习的中文短文本多标签分类模型

目前,中文短文本因其长度短、结构多样和缺乏上下文等特点,常规多标签分类算法无法对其有效区分。针对以上问题,论文提出一种基于深度学习的中文短文本多标签分类模型CRC-MHA。CRC-MHA模型在文本表示层摒弃常规使用Word2vec进行静态词嵌入的方式,采用BERT对输入句子进行动态词嵌入,借助海量预训练文本的优势更好地表征文本的上下文语义,同时在特征提取层设计了一种结合CNN、RCNN和多头自注意力机制的并行特征提取策略,加强捕捉文本内部的关键特征来提升多标签分类效果。实验结果表明,CRC-MHA模型在评价指标加权平均F1值上较BERT模型提高1.95%,较BERT-CNN模型提高0.42%,较BERT-RCNN模型提高0.34%,验证了模型的有效性。

基于词性标注的启发式在线日志解析方法

为了解决现有启发式日志解析方法中日志特征表示区分能力不足导致解析精度低、泛化差的问题,提出了一种启发式在线日志解析方法PosParser。该方法使用来源于触发词概念的功能词序列作为特征表示,包含解决复杂日志易过度解析问题的两阶段检测方法和处理变长参数日志的后处理流程。PosParser在16个真实日志数据集上取得了0.952的平均解析准确率,证明了功能词序列具有良好区分性、PosParser有良好的解析效果和鲁棒性。