原标题：译事 | 最新研究！机器翻译已对国际贸易产生重大影响
InScript: Narrative texts annotated with script information
This paper presents the InScript corpus (Narrative Texts Instantiating
Script structure). InScript is a corpus of 1,000 stories centered
around 10 different scenarios. Verbs and noun phrases are annotated with event and participant types, respectively. Additionally, the text
is annotated with coreference information. The corpus shows rich lexical variation and will serve as a unique resource for the study of
the role of script knowledge in natural language processing.
本文介绍了一种新的语料库，InScript (Narrative Texts Instantiating Script structure)。InScript 是一个由集中在10个不同场景的1000个故事组成的语料库。动词和名词短语分别被注解为事件和参与者类型。同时，文本由易混淆的信息注解。该语料库展现了词汇多样性，同时为研究自然语言处理中脚本知识的任务提供了独特的资源。
A group of researchers at the National Bureau of Economic Research (NBER) have written a working paper entitled "Does Machine Translation Affect International Trade? Evidence from a Large Digital Platform." The NBER is a private, non-profit organization that performs and disseminates economic research and one of the most influential think tanks on economic policy in the United States.
Is this word borrowed? An automatic approach to quantify the likeliness of borrowing in social media
Code-mixing or code-switching are the effortless phenomena of natural switching between two or more languages in a single conversation. Use of a foreign word in a language; however, does not necessarily mean that the speaker is code-switching because often languages borrow lexical items from other languages. If a word is borrowed, it becomes a part of the lexicon of a language; whereas, during code-switching, the speaker is aware that the conversation involves foreign words or phrases. Identifying whether a foreign word used by a bilingual speaker is due to borrowing or code-switching is a fundamental importance to theories of multilingualism, and an essential prerequisite towards the development of language and speech technologies for multilingual communities. In this paper, we present a series of novel computational methods to identify the borrowed likeliness of a word, based on the social media signals. We first propose context based clustering method to sample a set of candidate words from the social media data.Next, we propose three novel and similar metrics based on the usage of these words by the users in different tweets; these metrics were used to score and rank the candidate words indicating their borrowed likeliness. We compare these rankings with a ground truth ranking constructed through a human judgment experiment. The Spearman’s rank correlation between the two rankings (nearly 0.62 for all the three metric variants) is more than double the value (0.26) of the most competitive existing baseline reported in the literature. Some other striking observations are, (i) the correlation is higher for the ground truth data elicited from the younger participants (age less than 30) than that from the older participants, and (ii )those participants who use mixed-language for tweeting the least, provide the best signals of borrowing.
The paper explains that as one of the most significant technological advances today, artificial intelligence (AI) is poised to impact work, trade, and the economy. However, "empirical evidence documenting concrete economic effects of using AI is largely lacking." The study is one of the first to examine the material impact machine translation (MT), and AI in general, is having on trade and industry.
SyntaxNet Models for the CoNLL 2017 Shared Task
The paper by Erik Brynjolfsson, a Professor at MIT’s Sloan School of Management, Xiang Hui, and Meng Liu was published in August 2018. It looks at how eBay’s machine translation engines have affected international trade on the platform. This analysis is used as evidence "of direct causal links between AI adoption and economic activities."
We describe a baseline dependency parsing system for the CoNLL2017 Shared Task. This system, which we call “ParseySaurus,” uses the DRAGNN framework [Kong et al, 2017] to combine transition-based recurrent parsing and tagging with character-based word representations. On the v1.3 Universal Dependencies Treebanks, the new system outpeforms the publicly available, state-of-the-art “Parsey’s Cousins” models by 3.47% absolute Labeled Accuracy Score (LAS) across 52 treebanks.
该文由麻省理工学院斯隆管理学院教授Erik Brynjolfsson以及Xiang Hui和Meng Liu著成，于2018年8月发布。文章研究了eBay（译者注：美国最大的在线商品交易平台之一）的机器翻译引擎对其平台上的国际贸易有何影响。这个分析结果用于论证“人工智能应用和经济活动之间的直接因果联系”。
针对CoNLL2017给出了一个基线依赖解析系统。利用了DRAGNN framework [Kong et al, 2017]，结合基于转换的循环分析同时利用字符的词表示来标记。新系统性能很出色。
Ensemble of Neural Classifiers for Scoring Knowledge Base Triples
eBay has developed its own inhouse MT engines (eMT) for multiple language pairings. The study focuses on the English to Spanish eMT, which was introduced in 2014, and looks at its effect on international trade based on US exports.
This paper describes our approach for the triple scoring task at WSDM Cup 2017. The task aims to assign a relevance score for each pair of entities and their types in a knowledge base in order to enhance the ranking results in entity retrieval tasks. We propose an approach wherein the outputs of multiple neural network classifiers are combined using a supervised machine learning model. The experimental results show that our proposed method achieves the best performance in one out of three measures, and performs competitively in the other two measures.
According to then eBay Staff MT Language Specialist Juan Rowda in an article he wrote for Slator in November 2016, "human translation is not viable" for the company. He said as of that writing, the company had 800 million listings of around 300 words each. "It would take 1,000 translators 5 years to translate only the 60 million listings eligible for Russia," Rowda said, contextualizing the problem that eMT was meant to solve.
本论文描述了作者在2017 WSDM Cup（ACM网络搜索与数据挖掘国际会议，ACM International Conference on Web Search and Data Mining，简称WSDM）上所做的关于triple scoring工作的方法。这项工作要基于一个知识基准对每个实体对分配一个想关分数，以此来提高在实体检索工作中的排名结果。作者提出一个方法，利用一个监督的机器学习模型结合了多个神经网络分类器的输出。试验结果显示他们提出的方法在三钟测量中的一个得到了最后的结果，并且在其他两个中也具有显著的竞争性。
Improving Neural Machine Translation with Conditional Sequence Generative Adversarial Nets
The NBER paper finds that "the introduction of a machine translation system has significantly increased international trade on this platform, increasing exports by 17.5%." Although, using a number of different control samples, the results vary from a positive change of between 11.8% and 20.9%.
This paper proposes a new route for applying the generative adversarial nets (GANs) to NLP tasks (taking the neural machine translation as an instance) and the widespread perspective that GANs can’t work well in the NLP area turns out to be unreasonable. In this work, we build a conditional sequence generative adversarial net which comprises of two adversarial sub models, a generative model (generator) which translates the source sentence into the target sentence as the traditional NMT models do and a discriminative model (discriminator) which discriminates the machine-translated target sentence from the human-translated sentence. From the perspective of Turing test, the proposed model is to generate the translation which is indistinguishable from the human-translated one. Experiments show that the proposed model achieves significant improvements than the traditional NMT model. In Chinese-English translation tasks, we obtain up to 2.0 BLEU points improvement. To the best of our knowledge, this is the first time that the quantitative results about the application of GANs in the traditional NLP task is reported. Meanwhile, we present detailed strategies for GAN training. In addition, We find that the discriminator of the proposed model shows great capability in data cleaning.
eMT is having more of an impact on products that are differentiated, cheaper, or contain more words in listing titles, and is particularly effective for less experienced buyers, the study also found. Another interesting finding is that consumers seemed to "benefit from eMT more than sellers do […] because consumers gain both from reduced language frictions and also from lower prices."
Sparse Named Entity Classification using Factorization Machines
Named entity classification is the task of classifying text-based elements into various categories, including places, names, dates, times, and monetary values. A bottleneck in named entity classification, however, is the data problem of sparseness, because new named entities continually emerge, making it rather difficult to maintain a dictionary for named entity classification. Thus, in this paper, we address the problem of named entity classification using matrix factorization to overcome the problem of feature sparsity. Experimental results show that our proposed model, with fewer features and a smaller size, achieves competitive accuracy to state-of-the-art models.
Language Barriers Have Greatly Hindered Trade
The results in the research paper are consistent with other recent research, such as that of Johannes Lohmann in 2011 and Alejandro Molnar in 2013, who, according to the paper, "argue that language barriers may be far more trade-hindering than suggested by previous literature."
Encoding Sentences with Graph Convolutional Networks for Semantic Role Labeling
该研究论文中的结论与最近几年其它研究得出的结论一致，比如2011年Johannes Lohmann的研究和2013年Alejandro Molnar的研究，后者在论文中写道“语言障碍对贸易的制约可能比之前文献所述更甚。”
Semantic role labeling (SRL) is the task of identifying the predicate-argument structure of a sentence. It is typically regarded as an important step in the standard natural language processing pipeline, providing information to downstream tasks such as information extraction and question answering. As the semantic representations are closely related to syntactic ones, we exploit syntactic information in our model. We propose a version of graph convolutional networks (GCNs), a recent class of multilayer neural networks operating on graphs, suited to modeling syntactic dependency graphs. GCNs over syntactic dependency trees are used as sentence encoders, producing latent feature representations of words in a sentence and capturing information relevant to predicting the semantic representations. We observe that GCN layers are complementary to LSTM ones: when we stack both GCN and LSTM layers, we obtain a substantial improvement over an already state-of-the-art LSTM SRL model, resulting in the best reported scores on the standard benchmark (CoNLL-2009) both for Chinese and English.
eMT is not used for product categories, adverts, or other fixed areas of the webpages, but is specifically applied to searches. Buyers can search in Spanish, and eMT then enables eBay to match the search query to its listing titles, which are stored in English. From here, the eMT translates and displays the listing titles in Spanish, so the Spanish users only ever need to engage with Spanish content.
eBay’s previous translation solution for searches and item titles relied on Bing Translator, and the proprietary eMT system represents a "moderate quality upgrade", according to the paper.
FastQA: A Simple and Efficient Neural Architecture for Question Answering
Recent development of large-scale question answering (QA) datasets triggered a substantial amount of research into end-to-end neural architectures for QA. Increasingly complex systems have been conceived without comparison to a simpler neural baseline system that would justify their complexity. In this work, we propose a simple heuristic that guided the development of FastQA, an efficient end-to-end neural model for question answering that is very competitive with existing models. We further demonstrate, that an extended version (FastQAExt) achieves state-of-the-art results on recent benchmark datasets, namely SQuAD, NewsQA and MsMARCO, outperforming most existing models. However, we show that increasing the complexity of FastQA to FastQAExt does not yield any systematic improvements. We argue that the same holds true for most existing systems that are similar to FastQAExt. A manual analysis reveals that our proposed heuristic explains most predictions of our model, which indicates that modeling a simple heuristic is enough to achieve strong performance on extractive QA datasets. The overall strong performance of FastQA puts results of existing, more complex models into perspective.
The researchers also analyzed the impact of eMT in Europe, based on British and Irish exports to France, Italy and Spain, and concluded that "the rollout of eMT in Europe confirms the finding that machine translation increases exports on eBay."