Window Detection in Facades Using Heatmap Fusion

Journal of Computer Science and Technology - Window detection is a key component in many graphics and vision applications related to 3D city modeling and scene visualization. We present a novel...     

Skeleton parsing for complex question answering over knowledge bases

Answering complex questions involving multiple relations over knowledge bases is a challenging task. Many previous works rely on dependency parsing. However, errors in dependency parsing would influence their performance, in particular for long complex questions. In this paper, we propose a novel skeleton grammar to represent the high-level structure of a complex question. This lightweight formalism and its BERT-based parsing algorithm help to improve the downstream dependency parsing. To show the effectiveness of skeleton, we develop two question answering approaches: skeleton-based semantic parsing (called SSP) and skeleton-based information retrieval (called SIR). In SSP, skeleton helps to improve structured query generation. In SIR, skeleton helps to improve path ranking. Experimental results show that, thanks to skeletons, our approaches achieve state-of-the-art results on three datasets: LC-QuAD 1.0, GraphQuestions, and ComplexWebQuestions 1.1.     

基于ERNIE-BiGRU模型的摘要语步自动识别研究

学术文献的摘要是对文献主要内容的浓缩,摘要不同部分的语步具有不同的信息,语步的自动识别和抽取对于学术摘要的后续研究有着重要的应用价值,而目前语步识别的研究相对较少,并且相关算法的效果还需要提高。针对上述问题,该文提出了一种基于ERNIE-BiGRU模型的语步识别算法。该算法首先结合中文句法分析理论提出基于句法依存关系的多语步结构拆分法,对学术文献摘要多语步结构进行自动拆分,获得多个单语步结构;然后构建用于训练的单语步结构语料库,并利用知识增强语义表示预训练模型,训练出句子级词向量;最后将训练出的单语步结构词向量信息输入双向门限循环单元(BiGRU)进行摘要语步自动化识别,取得了良好的效果。实验结果表明,该算法具有较好的鲁棒性和较高的识别精度,在结构化和非结构化摘要上的识别准确率分别达到了96.57%和93.75%。     

基于桥连接的词典学习方法的语义解析

现阶段的语义解析方法大部分都基于组合语义,这类方法的核心就是词典。词典是词汇的集合,词汇定义了自然语言句子中词语到知识库本体中谓词的映射。语义解析一直面临着词典中词汇覆盖度不够的问题。针对此问题,该文在现有工作的基础上,提出了基于桥连接的词典学习方法,该方法能够在训练中自动引入新的词汇并加以学习,为了进一步提高新学习到的词汇的准确度,该文设计了新的词语—二元谓词的特征模板,并使用基于投票机制的核心词典获取方法。该文在两个公开数据集(WebQuestions和Free917)上进行了对比实验,实验结果表明,该文方法能够学习到新的词汇,提高词汇的覆盖度,进而提升语义解析系统的性能,特别是召回率。     

DiSeg 1.0: The first system for Spanish discourse segmentation

Nowadays discourse parsing is a very prominent research topic. However, there is not a discourse parser for Spanish texts. The first stage in order to develop this tool is discourse segmentation. In this work, we present DiSeg, the first discourse segmenter for Spanish, which uses the framework of Rhetorical Structure Theory and is based on lexical and syntactic rules. We describe the system and we evaluate its performance against a gold standard corpus, divided in a medical and a terminological subcorpus. We obtain promising results, which means that discourse segmentation is possible using shallow parsing.     

TSPN: Term-Based Semantic Peer-to-Peer Networks

In this paper,we propose Term-based Semantic Peerto-Peer Networks (TSPN) to achieve semantic search.For each peer,TSPN builds a full text index of its documents.Through the analysis of resources,TSPN o...     

An Empirical Approach to Spanish Anaphora Resolution

This paper documents the development of an empirically-basedsystem implemented in Prolog that automatically resolves severalkinds of anaphora in Spanish texts. These are pronominalreferences, surface-count anaphora, one-anaphora and ellipticalzero-subject constructions (i.e., sentences that omit theirpronominal subject). The resolution is based onrepresentations resulting from either partial or full parsing. Thesystem developed can also work on the output of a POStagger or with different dictionaries, without changing thegrammar. This grammar represents the syntactic information of eachlanguage by means of the Slot Unification Grammar formalism. The different kinds of information used for anaphora resolution in full and partial parsing are shown, as wellas evaluation results. The system has been adapted toEnglish texts, obtaining encouraging results that prove that itcan be applied with only a very few refinements to other languagesas well as Spanish and English. In addition, the differencesbetween English and Spanish anaphora are noted.     

海量COMTRADE数据并行解析算法

针对电力系统暂态监测数据向海量化发展的趋势，以及对各种串行编程分析软件造成的计算压力，研究了一种基于多核的、采用流水线设计思想的海量电力系统暂态数据交换通用格式（COMTRADE）数据文件并行解析算法。在面向二进制格式的并行解析算法中，采用了2阶段的流水线步骤:第1阶段为文件读取;第2阶段为多线程解析。在面向美国标准信息交换码（ASCII）格式的并行解析算法中，采用了3阶段的流水线步骤:前2个阶段与二进制格式的并行解析算法一样，第3阶段为块边缘解析，打破了传统的串行文本解析算法。算法通过有效利用以往被忽视的数据文件中的序号信息，大幅提升了并行解析的并行度。试验证明此算法可获得相当高的加速比，并且可随着CPU核数的增加和COMTRADE数据文件的加大获得线性加速比。