丁效, 宋凡, 秦兵, 等. 音乐领域典型事件抽取方法研究[J]. 中文信息学报, 2011, 25(2): 15-20.

Motivation

事件抽取是信息抽取研究中最具挑战性的任务之一,旨在把人们感兴趣的,用自然语言描述的事件 以结构化的形式呈现出来, 如什么人, 什么地方 , 什么时间 ,做了什么事。事件抽取在多文档文摘, 自动文摘,自动问答和信息检索领域有着广泛的应用。

目前的事件抽取方法采用信息抽取技术抽取预先定义的一种或者几种事件 ,然而不同领域的事件 类型互不相同,这样的方法依赖于人工定义的事件 类型,需要耗费大量的人工劳动 ,导致作为信息抽取 关键技术的事件抽取缺乏足够的自适应性。

Proposed method

本文借鉴 ACE 中事件抽取的相关概念, 并结合实际的需求做了相应的调整 , 将其转移到音乐领 域上来,从音乐新闻资讯中抽取出需要的结构化信 息。通过领域事件词聚类的方法自动发现音乐领域 典型事件 ,对典型事件分别从语料的获取 、标注 、事件的定义 、算法的应用都做了尝试 ,相同的方法可以平行的应用到其他的事件类型上。

音乐领域的事件抽取任务与ACE 事件抽取任 务大致相同,主要包括以下三个步骤:

1.事件类型发现:事件类型是事件抽取任务的一个基础,本文不同于传统事件抽取之处就在于不 是预先定义好事件类型体系 ,而是通过基于领域事 件词聚类的方法自动发现事件类型;

2.事件触发词及事件类别的识别:事件触发词是指引起事件发生的词 , 是决定事件类别的重要 特征;

3.事件元素的识别:事件的元素是指事件的参与者,本文为音乐领域的两个典型事件制定了模板, 模板的每个槽值对应着事件的元素。

Useful information

近些年来, 事件抽取一直吸引着许多研究机构和研究者的注意力。MUC (Message,Understanding Conference)会议(1987 ～ 1998)作为ACE(Automatic Content Ext raction)会议的前身 ,在上个世纪八 、九十年代对信息抽取领域起 到了 很 大 的促 进 作 用, 事件 抽 取 (Scenario Template)始终是这一会议的评测项目之一。ACE 也于 2005 年引入了事件抽取(Event Detection and Recognition 、Event Mention Detection)评测任务。

从技术 路线角度看 ,解决事件抽取问题的方法主要有两种: 基于模板的方法与基于机器学习的方法。基于模板 的主要通过手工或自动生成事件模板, 采用各种模式匹配算法将待抽取的句子和已经抽出的模板匹配。例如Yankova 的足球事件抽取系统以及 Lee的基于限定域 Onto logy 的气象事件抽取系统 等等。基于机器学习的方法把主要的精力放在分类器 的构建和特征的发现 、选择上 ,把事件抽取问题看成 分类问题 , 选择合适的特征使用分类器来完成。 Chieu 和Ng 于 2002 年首次在事件抽取中引入最大 熵分类器,用于事件抽取中事件元素的识别。

Hogenboom F, Frasincar F, Kaymak U, et al. A survey of event extraction methods from text for decision support systems[J]. Decision Support Systems, 2016, 85: 12-22.

Motivation

Event extraction, a specialized stream of information extraction rooted back into the 1980s, has greatly gained in popularity due to the advent of big data and the developments in the related fields of text mining and natural language processing. However, up to this date, an overview of this particular field remains elusive.

Proposed method

Therefore, we give a summarization of event extraction techniques for textual data, distinguishing between data-driven, knowledge-driven, and hybrid methods, and present a qualitative evaluation of these. Moreover, we discuss common decision support applications of event extraction from text corpora. Last, we elaborate on the evaluation of event extraction systems and identify current research issues.

Data-driven event extraction

Knowledge-driven event extraction

Hybrid event extraction

Research issues

1) the context-based advantage of data-driven, knowledge- driven, or hybrid approaches,

2) understanding the limitations of specific event extraction techniques,

3) the domain-dependency of event extraction procedures, affecting both their flexibility and effectiveness,

4) the scalability of event extraction approaches when dealing with big data,

5) the complexity of extracted events.

Some useful information

Since the early 2000s, there has been a notable shift from general information extraction from digital collections – extracting basic named entities like persons and orga- nizations – toward more advanced forms of text mining, including Event Extraction (EE) that requires the handling of textual content or data describing complex relations between entities [6].

Gottschalk S, Demidova E. Eventkg: A multilingual event-centric temporal knowledge graph[C]//European Semantic Web Conference. Springer, Cham, 2018: 272-287.

Motivation

One of the key requirements to facilitate semantic analy- tics of information regarding contemporary and historical events on the Web, in the news and in social media is the availability of reference know- ledge repositories containing comprehensive representations of events and temporal relations. Existing knowledge graphs, with popular examples including DBpedia, YAGO and Wikidata, focus mostly on entity-centric information and are insufficient in terms of their coverage and complete- ness with respect to events and temporal relations.

Proposed method

EventKG presented in this paper is a multilingual event-centric temporal knowledge graph that addresses this gap. EventKG incorporates over 690 thousand con- temporary and historical events and over 2.3 million temporal relations extracted from several large-scale knowledge graphs and semi-structured sources and makes them available through a canonical representation.

Van Hage W R, Malaisé V, Segers R, et al. Design and use of the Simple Event Model (SEM)[J]. Web Semantics: Science, Services and Agents on the World Wide Web, 2011, 9(2): 128-136.

Motivation

Events have become central elements in the representation of data from domains such as history, cultural heritage, multimedia and geography.

Proposed method

The Simple Event Model (SEM) is created to model events in these various domains, without making assumptions about the domain-specific vocabularies used. SEM is designed with a minimum of semantic commitment to guarantee maximal interoperability.

In this paper, we discuss the general requirements of an event model for Web data and give examples from two use cases: historic events and events in the maritime safety and security domain. The advantages and disadvantages of several existing event models are discussed in the context of the historic example. We discuss the design decisions underlying SEM. SEM is coupled with a Prolog API that enables users to create instances of events without going into the details of the implementation of the model. By a tight coupling to existing Prolog packages, the API facilitates easy integration of event instances to Linked Open Data. We illustrate use of the API with examples from the maritime domain.

SEM’s classes are divided in three groups: core classes, types, and constraints. This is illustrated in Fig. 1. There are four core clas- ses: sem:Event (what happens), sem:Actor (who or what partici- pated), sem:Place (where), sem:Time (when). Each core class5 has an associated sem:Type class, which contains resources that indicate the type of a core individual. Individuals and their types are usually borrowed from other vocabularies. For example, the sem:Place ‘‘Indo- nesia’’ (tgn:1000116) from Fig. 4 and its sem:PlaceType ‘‘republic’’ (tgn:82171) are borrowed from the Getty Institute’s Thesaurus of Geographical Names (TGN).

Wojtowicz J, Wallace W A. Traffic management for planned special events using traffic microsimulation modeling and tabletop exercises[J]. Journal of Transportation Safety & Security, 2010, 2(2): 102-121.

Proposed method

This article describes the use of microsimulation for traffic management of special events, including evacuations and traffic incidents.

Specifically, the research employed tabletop exercises using the microsimulation software TransModeler to present traffic management alternatives for special events to stakeholders.

The exercises involved analysis of evacuation or egress conditions with current traffic management plans, review of possible high-impact incident scenarios, and testing and formulation of improvements to current traffic management procedures, including incident response.

Two case studies were developed for the tabletop exercises; one the evacuation of a minor-league baseball stadium and another concerned with vehicular egress plans for a large urban indoor arena. Both venues were in New York State’s Capital District region. The results demonstrated that the real-time use ofmicrosimulation during a tabletop exercise elicited many new traffic management strategies. Also, a cost–benefit analysis of using microsimulation models in conjunction with tabletop exercises for these types of events was explored. The project was made possible by funding from the New York State Department ofTransportation and the Federal Highway Administration. Keywords

Socher R, Chen D, Manning C D, et al. Reasoning with neural tensor networks for knowledge base completion[C]//Advances in neural information processing systems. 2013: 926-934.

Motivation

Knowledge bases are an important resource for question answering and other tasks but often suffer from incompleteness and lack of ability to reason over their discrete entities and relationships.

not all common knowledge that is obvious to people is expressed in text [5, 6, 2, 7]. We adopt here the complementary goal of predicting the likely truth of additional facts based on existing facts in the knowledge base. Such factual, common sense reasoning is available and useful to people. For instance, when told that a new species of monkeys has been discovered, a person does not need to find textual evidence to know that this new monkey, too, will have legs (a meronymic relationship inferred due to a hyponymic relation to monkeys in general).

Proposed method

In this paper we introduce an expressive neural tensor network suitable for reasoning over relationships between two entities.

Previous work represented entities as either discrete atomic units or with a single entity vector representation. We show that performance can be improved when entities are represented as an average of their constituting word vectors. This allows sharing of statistical strength between, for instance, facts involving the “Sumatran tiger” and “Bengal tiger.”

Lastly, we demonstrate that all models improve when these word vectors are initialized with vectors learned from unsupervised large corpora. We assess the model by considering the problem of predicting additional true relations between entities given a subset of the knowledge base. Our model outperforms previous models and can classify unseen relationships in WordNet and FreeBase with an accuracy of 86.2% and 90.0%, respectively.

Contribution

The first contribution of this paper is the new neural tensor network (NTN), which generalizes several previous neural network models and provides a more powerful way to model relational information than a standard neural network layer.

The second contribution is to introduce a newway to represent entities in knowledge bases. Previous work [8, 9, 10] represents each entity with one vector. However, does not allow the sharing of statistical strength if entity names share similar substrings. Instead, we represent each entity as the average of its word vectors, allowing the sharing of statistical strength between the words describing each entity e.g., Bank of China and China.

The third contribution is the incorporation of word vectors which are trained on large unlabeled text. This readily available resource enables all models to more accurately predict relationships.

Yeoh B S A, Huang S. Introduction: fluidity and friction in talent migration[J]. Journal of Ethnic and Migration Studies, 2011, 37(5): 681-690.

Research work

We draw out two key themes that inform the arguments advanced by the papers in this special issue.

First, instead of hyper-fluid global nomads who move effortlessly across frictionless global space, the papers show that moving and belonging are not simply motivated by economic logic but are folded into broader social, cultural and political considerations and conditioned by the power geometries ofrace, nationality and gender.

Second, different groups of talent migrants are attracted to and positioned differently in specific spaces of encounter within cities, generating a politics of identity that influences everyday life.

Taken together, these two themes highlight the need to avoid artificially bifurcating the lives of talent migrants into the economic and social/cultural spheres, and instead to argue for interweaving them within the same frame in understanding the mobility of the highly skilled.

The Cultural Politics of Moving and Belonging

The Cultural Politics of Place and Identity

Useful information

In more recent times, however, the emerging scholarship on highly skilled economic migrants has attempted to move beyond assumptions of hyper-mobility and cosmopolitan sophistication to portray them as embodied bearers of culture, ethnicity, class or gender.

A second related move in the emerging literature is to counter earlier assumptions that transnational elites are perpetually ‘rootless merchant sojourners’ (Cheah 2001: 135), ‘cosmopolitans’ who are ‘basically indifferent to where they lived’,or ‘cosmopolites’ who are ‘habitants of a vast universe’ (Robbins 1998: 3);

Biao X. A ritual economy of ‘talent’: China and overseas Chinese professionals[J]. Journal of Ethnic and Migration Studies, 2011, 37(5): 821-838.

Motivation

Since the Guangzhou municipality in south China organised the first Overseas Students Fair in 1998, large conventions aimed at recruiting overseas Chinese professionals (OCPs) have become a regular scene in major cities in China. These conventions constitute one of the most visible means of the Chinese government’s engagement with the 800,000 OCPs who remain overseas after receiving tertiary education abroad. Characteristic of the conventions, and OCP policies in general, is a highly ‘materialistic’ thinking: it is argued that OCPs deserve generous financial rewards because they are economically and technologically beneficial to China, and that financial reward is the most feasible means to attract them back.

My ethnographic data, however, reveal that the language ofeconomism is communicated in a highly ritualistic manner and, conversely, political rituals serve as a crucial part ofthe conventions. The ritualised economic- and technological-determinist discourse appears apolitical, yet acquires strong mobilising and legitimating power, and is thus particularly effective in accommodating OCPs into the established political order. The concept ‘ritual economy’ denotes such deep intertwining between the economic, the ritualistic and the political.

The Making of ‘haiwai xuezi’

Central Government: One Hand Soft, One Hand Hard

Local Government: Government Sets the Stage, Businesses Run the Opera

Latukha M, Nintuona Soyiri J. Determinants of Talent Mobility in Africa: Talent Attraction and Retention Practices in Ghana[J]. 2018.

Motivation

Due to existing globalization processes, African countries need to have more extensive economic growth and development to be better integrated to the global environment. Knowledge transfer from developed to developing countries can be considered as essential tools, whereas talents from expatriates may serve as knowledge acquisition and assimilation mechanism (Artuc et al.2015).Other researchers including (Allan,2015;Bimbassis,2012) have all expressed their conviction that an inflow of skilled expatriates into Africa is necessary to enhance Africa’s economic growth. This is very necessary for example to enable many African countries tap into their diverse natural resources, promote economic growth and reverse the position of Africa from being the second largest continent, but the poorest in the world.(Bimbassis,2012)

Research work

The paper investigates the determinants of talent mobility in Africa and the role of talent management (TM) practices in talent mobility to Ghana, paying special attention to the determinants of talent mobility in African countries aiming to enhance more extensive economic development by knowledge transfer by expatriates from developed countries. A survey will be conducted on 250 -300 expatriates and 50 CEOs and 20 TM consultants already working in Ghana. We expect the results to confirm that with an efficient TM system supported by the government, firms in Ghana can convert expatriates to migrants to enhance expatriate knowledge transfer into Africa, particularly Ghana.

Research goal

the research goal is to define the factors that influence talent migration to Ghana and to investigate the role of talent management practices in attracting and retaining expatriates in Ghana.

Specifically, how talent management practices should be developed and implemented to be able to convert “expatriates to migrants” for efficient knowledge transfer to Ghana?

Research question

  1. What are the determinants of expatriate talent mobility from developed countries to Africa, particularly to Ghana?
  2. What talent management practices can be used in Ghana context to attract and retain expatriates and stimulate knowledge transfer from developed countries?
  3. How can Ghana use talent management practices to neutralize the negative perception of Ghana as a destination country for expatriates and covert them to migrants?

Exploring the global mobility phenomenon

Global mobility which is defined by some scholars as the movement of both skilled and unskilled people locally and internationally (Altbach & Bassett, 2004) is a phenomenon that is now being strategically exploited and managed by stake holders as a developmental tool.

Useful information

The role of global migration in expatriate knowledge transfer is so strategic because it serves as a supply base from which expatriates who are generally known as mechanisms for knowledge transfer are distributed to both developed and developing countries. (Cascio and Aguinis 2007) have noted that the successful MNE of the future will be required to source human and intellectual resources regardless of their global location, owing to the increasing competition.

Al Ariss A, Sidani Y. Comparative international human resource management: Future research directions[J]. Human Resource Management Review, 2016, 26(4): 352-358.

Motivation

There are numerous under-researched topics that fall under the umbrella of comparative international human resource management (IHRM) theory and practice. We address three in particular: talent management (TM), international mobility, and diversity. The degree of convergence of HR practices with Western-style so-called “best practices” is a topic of much dispute. We argue that sociocultural and institutional factors are critical determinants of HR practices, and we underscore the importance of the sociocultural context in IHRM. The reality unfolding is that of multiple factors, not exclusive to international best practices, playing a role in the development of HR practices in various international contexts.

Convergence, divergence, and crossvergence theories

The convergence perspective posits that as societies become more similar to one another in terms of industrialization and the use of technology, values will eventually converge towards Western capitalism given that this is where most industrialization has traditionally occurred (Ralston, 2008).

Earlier research in cross-cultural theory (Hofstede, Hofstede, & Minkov, 1991; House, Hanges, Javidan, Dorfman, & Gupta, 2004), and the ongoing search for cross-cultural variations in value structures (e.g., Schwartz & Bilsky, 1990) strongly indicate that–despite technological advancements, globalization forces, and the growing powers of MNCs–people from different cultural and geographical contexts are not the same and HR systems dealing with them are not, and need not be, the same. The established scholarship in international business (IB) about localization strategies, for example, suggests that companies are aware of the need to establish varying competitive strategies to fit different contexts and consumer tastes and behaviors. For in- ternal customers–the employees of the firm, one can assume that the same logic applies.

The crossvergence perspective, as far as HR practices are concerned, has been under-studied in HR scholarship. It promotes a ‘best fit’ approach to IHRM. Earlier HR research hints at crossvergence in certain circumstances. Sidani and Al Ariss (2014),for example, suggest that MNCs operate in such a way that certain practices converge (given their global usage) while other practices diverge (given local contexts), thus offering a crossvergence perspective.

Talent management and IHRM

The numerous definitions ascribed to the term talent management (TM) highlight the progress that has been made in the field of modern HR. In this section, we build on the recent literature review proposed by Al Ariss, Cascio, and Paauwe (2014).

Initially, the essence of TM was often conceived of as recruitment of the highest calibre of individuals, notably at managerial level, and identification of the particular qualities that distinguish a successful manager.

Two different conceptions of TM are acknowledged: exclusive and inclusive. The former focuses exclusively on a top elite of employees, while the latter applies to all employees. In terms of IHRM, globalization has prompted the development of a more internationally focused understanding of TM, known as global talent management (GTM).

GTM is interpreted by Vaiman, Scullion, and Collings (2012) as comprising the various means by which an organization achieves the attraction, selection, development, and retention of top talent in the highest global positions. Each company’s particular set of circumstances requires uniquely appropriate TM practices. The identification of core precepts of GTM was attempted by Stahl et al. (2012) for organizations to adopt and implement in attaining optimal performance. They find that companies would do well to design their TM strategies based on their own company values and objectives rather than seek to replicate successful TM practices used by other organizations. Therefore GTM is best guided by a “best fit” approach, rather than that of “best practice”.

Schuler, Jackson, and Tarique (2011) and Vaiman et al. (2012) identify several factors and forces that impact the TM process. We argue that the configuration of those forces and how they impact companies in different national contexts will yield TM practices that vary. For example, talent shortages internationally necessitate that companies look at talent, not only from a national perspective, but also from a global one. While the impact of talent shortages means that companies will converge on their need to manage global talent, it does not necessitate that they converge on how such talent is attracted or retained. Some MNCs short of talent will emphasize grooming internal talent; others would rather adopt competitive strategies to attract external ones.

International migration and IHRM

International migration has been on the rise over the last half century. Such global movement is attributed mainly to demo- graphic changes, lack of skilled workers globally, globalization of the economy and job market, and easier access across geographic as well as institutional and cultural borders (Howe-Walsh & Schyns, 2010).

Diversity and IHRM

Divergence of HR practices is influenced, as per the lens of institutional theory, to different degrees by the pressures imposed by the powers that be. Laws and regulations enforced by these powers will coerce an organization to toe a particular line as far as its HR practices are concerned.

Agrawal A, Kapur D, McHale J, et al. Brain drain or brain bank? The impact of skilled emigration on poor-country innovation[J]. Journal of Urban Economics, 2011, 69(1): 43-55.

Motivation

The development impact of skilled migration from poor countries has long been a contentious issue. Scholars are even far from a consensus on the narrower question: What is the impact on inno- vation when a poor country loses a large fraction of its science and engineering workforce through emigration?

The development prospects of a poor country or region depend in part on its capacity for innovation. In turn, the productivity of its innovators, whom are often concentrated around urban centers, depends on their access to technological knowledge. The emigration of highly skilled individuals weakens local knowledge networks (brain drain) but may also help remaining innovators access valuable knowledge accumulated abroad (brain bank).

Research framework

We develop a model in which the size of the optimal innovator Diaspora depends on the competing strengths of co-location and Diaspora effects for accessing knowledge. Then, using patent citation data associated with inventions from India, we estimate the key co-location and Diaspora parameters. The net effect of innovator emigration is to harm domestic knowledge access, on average. However, knowledge access conferred by the Diaspora is particularly valuable in the production of India’s most important inventions as measured by citations received. Thus, our findings imply that the optimal emigration level may depend, at least partly, on the relative value resulting from the most cited compared to average inventions.

Wang, Zhen, et al. “Knowledge graph embedding by translating on hyperplanes.” Twenty-Eighth AAAI conference on artificial intelligence. 2014.

Motivation

We deal with embedding a large scale knowledge graph composed of entities and relations into a continuous vector space.

TransE is a promising method proposed recently, which is very efficient while achieving state-of-the-art predictive per- formance. We discuss some mapping properties of relations which should be considered in embedding, such as reflex- ive, one-to-many, many-to-one, and many-to-many. We note that TransE does not do well in dealing with these proper- ties. Some complex models are capable of preserving these mapping properties but sacrifice efficiency in the process.

Research work

To make a good trade-off between model capacity and efficiency, in this paper we propose TransH which models a relation as a hyperplane together with a translation operation on it. In this way, we can well preserve the above mapping properties of relations with almost the same model complexity of TransE. Additionally, as a practical knowledge graph is often far from completed, how to construct negative examples to reduce false negative labels in training is very important. Utilizing the one-to-many/many-to-one mapping property of a relation, we propose a simple trick to reduce the possibility of false negative labeling. We conduct extensive experiments on link prediction, triplet classification and fact extraction on bench- mark datasets like WordNet and Freebase. Experiments show TransH delivers significant improvements over TransE on predictive accuracy with comparable capability to scale up.