机器翻译译文质量估计综述

机器翻译译文质量估计(Quality Estimation,QE)是指在不需要人工参考译文的条件下,估计机器翻译系统产生的译文的质量,对机器翻译研究和应用具有很重要的价值。机器翻译译文质量估计经过最近几年的发展,取得了丰富的研究成果。该文首先介绍了机器翻译译文质量估计的背景与意义;然后详细介绍了句子级QE、单词级QE、文档级QE的具体任务目标、评价指标等内容,进一步概括了QE方法发展的三个阶段: 基于特征工程和机器学习的QE方法阶段,基于深度学习的QE方法阶段,融入预训练模型的QE方法阶段,并介绍了每一阶段中的代表性研究工作;最后分析了目前的研究现状及不足,并对未来QE方法的研究及发展方向进行了展望。     

面向小语种机器翻译的平行语料库构建方法

神经机器翻译模型的训练效果在很大程度上取决于平行语料库的规模和质量。除了一些常见语言外,汉语与小语种间高质量平行语料库的建设一直处于滞后状态。现有小语种平行语料库多采用自动句子对齐技术利用网络资源构建而成,在文本质量和领域等方面有诸多局限性。采用人工翻译的方式可以构建高质量平行语料库,但是缺乏相关经验和方法。文中从机器翻译实践者和研究者角度出发,介绍了经济高效的人工构建小语种平行语料库的工作,包括其总体目标、实施过程、流程细节和最后结果。在构建过程中尝试并积累了各种经验,形成了小语种到汉语平行语料库构建方法、建议的总结。最终,成功构建了波斯语到汉语、印地语到汉语、印度尼西亚语到汉语各50万条高质量平行语料。实验结果表明,所构建的平行语料库有较好的质量,提高了小语种神经机器翻译模型的训练效果。     

火花放电原子发射光谱类型标准化校正方式的探讨

在国内,火花放电原子发射光谱分析广泛使用类型标准化进行方法校正。在国外,标准和文献中鲜见使用此方法的相关论述。类型标准化主要采用平移校正和转动校正两种方式,哪种方式更加合理也鲜见报道。国外设备类型标准化的默认设置优先采用转动校正方式,相关国内标准对最优校正方式的规定尚不明确,此默认设置的合理性有待探讨。实验选用低合金钢20CrNi2Mo、R407标准样品和不锈钢317L、0Cr18Ni9标准样品,引用国内相关标准,以正确度临界差为评判依据,模拟类型标准化样品和待测样品“十分接近”和“接近”两种情况下平移校正和转动校正的数据正确度。经数据统计分析,平移校正分析结果均满足要求,转动校正结果在“接近”情况下部分元素不满足要求。结合相关国家标准中元素含量范围和精密度数据进行分析,通过计算允许最大偏倚量并制作曲线图方式展开分析,得出如下结论:在满足文中类型标准化控制要点前提下,分析设置更适合于采用平移校正方式。     

带权决策表的属性约简

属性约简是粗糙集理论的重要应用。考虑将决策表中的每行都作为一条决策规则时，若把表中出现相同决策规则的次数作为权，可得到带权决策表。提出了关于带权决策表的正域约简相应的辨识矩阵并给出了证明，从而得到了约简算法。相比于决策表中的正域约简时发现，通过将决策表转化为带权决策表后，再利用算法1进行约简时，其在一定程度上优于前者。提出了近似分类精度约简相应的辨识矩阵并给出了证明。对于2个算法，在选取的UCI数据集上进行了实验验证。通过实验进一步说明了所提出算法的可行性和有效性。     

基于循环一致性对抗网络的室内火焰图像场景迁移

基于深度学习的视频火灾探测模型的训练依赖于大量的正负样本数据,即火灾视频和带有干扰的场景视频。由于很多室内场合禁止点火,导致该场景下的火灾视频样本不足。本文基于生成对抗网络,将其他相似场景下录制的火焰迁移到指定场景,以此增广限制性场合下的火灾视频数据。文中提出将火焰内核预先植入场景使之具备完整的内容信息,再通过添加烟雾和地面反射等风格信息,完成场景与火焰的融合。该方法克服了现有多模态图像转换方法在图像转换过程中因丢失信息而造成的背景失真问题。同时为减少数据采集工作量,采用循环一致性生成对抗网络以解除训练图像必须严格匹配的限制。实验表明,与现有多模态图像转换相比,本文方法可以保证场景中火焰形态的多样性,迁移后的场景具有较高的视觉真实性,所得结果的FID与LPIPS值最小,分别为119.6和0.134 2。     

基于聚类的数据加权优化在犯罪预测中的应用

近年来，我国传统暴力犯罪与成年人犯罪呈下降态势，但是，犯罪案由层出不穷。为有效提升公安实践工作中犯罪预测能力，打击各类违法犯罪事件,本文针对犯罪数据，提出一种新型犯罪预测模型。利用密度聚类分析方法将犯罪数据分类，然后进行数据降维提取关键属性生成特征数据，继而对特征数据进行加权优化并采用机器学习的方式对特征数据进行学习，从而预测犯罪案由。实验结果表明，与传统方法相比，本文方法具有更好的预测效果，为公安实践工作中类似案件的侦破和预防，提供新的路径支撑。     

Policy translation and configuration using dynamic template

To solve the problem of complex,cumbersome and error-prone configuration of security devices caused by the heterogeneous configuration modes in complex networks,a dynamic template-based scheme for translating and configuring policy was proposed.In detail,considering the code’s features,the code-based template for translating policies was constructed to configure the command line conversion,and the keyword comparison method was used to ensure the accuracy of policy configuration.Experiments show that the scalability and the accuracy of the proposed scheme.     

Multicriteria decision-making using Archimedean aggregation operators in Pythagorean hesitant fuzzy environment

In multicriteria decision-making (MCDM), the existing aggregation operators are mostly based on algebraic t-conorm and t-norm. But, Archimedean t-conorms and t-norms are the generalized forms of t-conorms and t-norms which include algebraic, Einstein, Hamacher, Frank, and other types of t-conorms and t-norms. From that view point, in this paper the concepts of Archimedean t-conorm and t-norm are introduced to aggregate Pythagorean hesitant fuzzy information. Some new operational laws for Pythagorean hesitant fuzzy numbers based on Archimedean t-conorm and t-norm have been proposed. Using those operational laws, Archimedean t-conorm and t-norm-based Pythagorean hesitant fuzzy weighted averaging operator and weighted geometric operator are developed. Some of their desirable properties have also been investigated. Afterwards, these operators are applied to solve MCDM problems in Pythagorean hesitant fuzzy environment. The developed Archimedean aggregation operators are also applicable in Pythagorean fuzzy contexts also. To demonstrate the validity, practicality, and effectiveness of the proposed method, a practical problem is considered, solved, and compared with other existing method.     

An AW-HARIS Based Automated Segmentation of Human Liver Using CT Images

In the digestion of amino acids, carbohydrates, and lipids, as well as protein synthesis from the consumed food, the liver has many diverse responsibilities and functions that are to be performed. Liver disease may impact the hormonal and nutritional balance in the human body. The earlier diagnosis of such critical conditions may help to treat the patient effectively. A computationally efficient AW-HARIS algorithm is used in this paper to perform automated segmentation of CT scan images to identify abnormalities in the human liver. The proposed approach can recognize the abnormalities with better accuracy without training, unlike in supervisory procedures requiring considerable computational efforts for training. In the earlier stages, the CT images are pre-processed through an Adaptive Multiscale Data Condensation Kernel to normalize the underlying noise and enhance the image’s contrast for better segmentation. Then, the preliminary phase’s outcome is being fed as the input for the Anisotropic Weighted–-Heuristic Algorithm for Real-time Image Segmentation algorithm that uses texture-related information, which has resulted in precise outcome with acceptable computational latency when compared to that of its counterparts. It is observed that the proposed approach has outperformed in the majority of the cases with an accuracy of 78%. The smart diagnosis approach would help the medical staff accurately predict the abnormality and disease progression in earlier ailment stages.