共查询到20条相似文献，搜索用时 15 毫秒
1.
《Information processing & management》2023,60(4):103153
Few-Shot Event Classification (FSEC) aims at assigning event labels to unlabeled sentences when limited annotated samples are available. Existing works mainly focus on using meta-learning to overcome the low-resource problem that still requires abundant held-out classes for model learning and selection. Thus we propose to deal with the low-resource problem by utilizing prompts. Further, existing methods suffer from severe trigger biases that may result in ignorance of the context. That is, the correct classifications are gained by looking at only the triggers, which hurts the model’s generalization ability. Thus, we propose a knowledgeable augmented-trigger prompt FSEC framework (AugPrompt), which can overcome the bias issues and alleviates the classification bottleneck brought by insufficient data. In detail, we first design an External Knowledge Injection (EKI) module to incorporate an external knowledge base (Related Words) for trigger augmentation. Then, we propose an Event Prompt Generation (EPG) module to generate appropriate discrete prompts for initializing the continuous prompts. After that, we propose an Event Prompt Tuning (EPT) module to automatically search prompts in the continuous space for FSEC and finally predict the corresponding event types of the inputs. We conduct extensive experiments on two public English datasets for FSEC, i.e., FewEvent and RAMS. The experimental results show the superiority of our proposal over the competitive baselines, where the maximum accuracy increase compared to the strongest baseline reaches 10.8%. 相似文献
2.
The Internet is becoming an increasingly important and pervasive channel to market for many organisations. Despite its importance, and the continued pressure to justify IT expenditure, few organisations undertake comprehensive channel evaluation. Market leading evaluation firms provide technical and operational metrics for their clients such as the number of hits per page and site response times, but more sophisticated concepts such as user value and long-term business benefits remain underexplored in practice. In contrast, there is a growing academic literature on channel evaluation. Many frameworks and metrics have been proposed recently. This paper brings theory and practice together by synthesizing existing frameworks proposed by academics with those used by the market leaders in Internet channel evaluation. The resulting framework has two levels—a business- and a user-level. The framework is validated by Internet consultants, channel managers and channel users in three different sectors: retail, financial services and higher education. The framework is refined following the validation in response to the need for a simpler, more usable set of metrics. The outcome is a framework split into three “sets”. Set A constitutes the foundation stone of an Internet channel evaluation programme and consists of a core set of objective user-level metrics. Set B contains a further set of more sophisticated user-level metrics. Set C addresses business-level metrics, which enable the long-term contribution of the Internet channel to be evaluated. 相似文献
3.
4.
Sponsored search is an online advertising channel that has gained momentum worldwide. The key challenge is deciding on the types of keywords to bid on and matching options to utilise for the keywords. In this paper, we address this problem by providing a broad analysis on how the various traffic search metrics (length, CTR, average cost per click (CPC), average position, and quality score) influence the bidding results as the keyword matching option becomes broader, that is, from exact, to phrase to broad. Drawing on the shopping goals theory, we also establish the profile of the metrics associated with a more focused search intent across the matching options. Using a random sample of keywords selected from 9 640 keywords that online advertisers have bid on, spanning a variety of markets, the results indicate that as the matching option becomes narrower, that is, from broad to exact, the keyword traffic metrics increase in general, except for cost, which does not differ significantly across the matching options. Longer keywords, which are typically associated with a more focussed search intent, generate more clicks and have a higher quality measure on average across all matching options. The longer keywords are cheaper for the exact match option, but more expensive for the other matching options. The results are inconclusive with regard to the position that longer keywords occupy on a results page across the matching options. Thus, the narrower matching options and longer keywords matching those that a customer would typically use to search for a company's goods and services need to be targeted to ensure higher visits to a company's website. 相似文献
5.
基于物联网和数据中台技术的自然资源要素综合观测平台构建 总被引：1，自引：0，他引：1
自然资源要素综合观测数据具有来源多、类型复杂和数据量庞大等特点,需要建立自然资源要素综合观测一体化平台对数据进行统一管理和利用。本文基于物联网与数据中台技术,构建了自然资源要素综合观测一体化平台,主要包括自然资源要素数据中心、物联网中心和模型云中心三大模块,实现了从传感器终端到数据中台、模型云的自动数据传输和互联共享,以及数值分析、加工和综合应用,达到了可视化效果。为自然资源统一管理、全国重要生态系统保护和修复重大工程规划提供技术支撑。 相似文献
6.
Information technology (IT) engagement is defined as a need to spend more time using IT. Practice-based examples show that IT engagement can have adverse effects in organizations. Although users can potentially get more work done through IT engagement, observations show that the users might jeopardize their well-being and hamper their work performance. We aimed to investigate this complexity in the research on IT engagement by examining its potential antecedents and outcomes in organizations. Considering the potentially mixed outcomes, we developed a model to examine the effects of IT engagement on personal productivity and strain. We also aimed to explain the antecedents of IT engagement by drawing on the collective expectations for IT use. In particular, we examined the extent to which normative pressure on IT use drives users’ information load and IT engagement. Finally, we sought to understand whether users’ attempts to avert dependency on IT use reduced their IT engagement. Several hypotheses were developed and tested with survey data of 1091 organizational IT users. The findings help explain the role of normative pressure as a key driver of IT engagement and validate the positive and negative outcomes of IT engagement in organizations. 相似文献
7.
地理信息系统在矿产资源勘查领域中的应用 总被引：1，自引：0，他引：1
地理信息系统(GIS)是近年发展起来的综合处理和分析空间数据的一种技术系统.与现代地球及其相关科学日益增长的需求相适应,以处理地球上任何具有空间方位的海量信息为特征,具定量、定时、定位等优点,近10年来已在地质矿产勘查中得到广泛应用.一个区域各种地质资料(图形、图像、文字、逻辑、数值)的GIS分析实际上代表该区域现阶段较为客观的总认识.目前,野外收集资料、数据建库、GIS分析等尚存在规范化、标准化等问题,GIS本身解决诸多专业性较强地质问题的能力亦不足.但GIS的进一步发展与完善必将使地质矿产勘查进入一个数字化的新时期. 相似文献
8.
In the literature on new product development, most existing studies on the end of the design process concentrate on managing ramp-up in the field of manufactured products. This situation poses a problem at a time when our economies increasingly depend upon services and products are more and more related with sophisticated services that provide value for customers and producers. This article examines the management of the final phases of the design process of an innovative product-related service. Our research thus makes three contributions: (1) an analysis of the implementation process shows that the simultaneity of the production and consumption of a service means that three types of learning - technical, sales and uses - take place at the same time. Launch management strategies have thus to be adapted; (2) an analysis of the data collected confirms this difference by bringing to light great contrasts in these different aspects of learning; (3) this led us to identify a field that needs exploration by researchers in product and service innovation: the design of the sales process. Furthermore we underline two scenarios to manage the launch of innovative product-related services. 相似文献
9.
网络强化组织正面临开发与使用有效的衡量标准以评价电子商务活动成功的压力.针对电子商务应用成功衡量标准的设计问题,发现了新衡量标准的五个基本特征,设计了全面而合理的电子商务应用成功衡量标准的三个关键维度,并识别了这些关键衡量标准之间的逻辑关系.对BS公司的案例分析验证了这三个维度衡量标准的有效性. 相似文献
10.
Organizations considering green IT initiatives seek to reconcile two significant goals: environmental sustainability (“doing good”) and business profitability (“doing well”). These two purposes, however, are not necessarily aligned and can engender a dilemma in which investing to achieve environmental sustainability may threaten business profitability to some extent. Thus, this research builds on the theoretical perspective of corporate ecological responsiveness and proposes three types of strategic drivers for green IT initiatives (economic, authority, and moral drivers). How various types of organizations may be motivated by these green IT initiative drivers will affect whether organizations can reconcile the objectives of environmental sustainability and business profitability. We further deduce findings from multiple case studies of eight organizations in China and Singapore. The findings illustrate that the characteristics of any given organization influence why various driving factors affect their green IT initiative decisions. Additionally, “doing good” and “doing well” are reconciled when organizations consider the short-term investment and long-term benefit, design appropriate strategy, and receive severe external pressure. We also discuss contributions to research and practice. 相似文献
11.
To reduce information technology (IT) development costs, more firms have begun to outsource IT-related activities by partnering with IT vendors. As knowledge is a valuable asset in IT development, knowledge sharing between vendors and business clients becomes critical. However, the motivation behind IT vendors’ willingness to share knowledge with client firms is not sufficiently understood. To shed light on the nature of knowledge sharing within vendor–client partnerships, we examine the influence of performance feedback and managerial mindset on vendors’ motivation to share knowledge with their clients. We adopt a multi-method approach involving both a scenario-based field experiment with 164 vendor managers (Study 1) and a field survey of 112 vendor managers involved in IT development (Study 2). We find that when vendors’ performance exceeds their aspiration levels, they are motivated to share knowledge with clients. Such motivational effects are more pronounced for vendor managers exhibiting abstract mindsets. Our study is of significant value to researchers and practitioners, affording both groups a keener, deeper, and more robust appreciation for how knowledge sharing in vendor–client partnerships can be managed more effectively. 相似文献
12.
IT已经成为企业创造增值产品和服务的基础性平台。随着企业对信息、系统、通讯等信息及相关技术的日益依赖和随之而来的大量IT投资的同时,企业又面临着IT投资和运用过程的威胁和商业风险的规避问题。研究以IT运用为中间变量,构建IT治理与IT绩效之间的关系以及IT治理对IT绩效影响的理论模型,并用结构方程模型进行实证分析。结果表明,IT治理与IT绩效之间没有积极的直接关系,但通过IT运用中介作用,IT治理对IT绩效的各个维度具有积极的正向影响,因此IT治理是通过促进IT运用水平的提升从而获得相应的IT投资效果。该研究结果进一步完善了IT治理与组织绩效之间的相关理论,并对信息化建设路径的选择具有一定的实践意义。 相似文献
13.
Matheus A. Marins Felipe M.L. Ribeiro Sergio L. Netto Eduardo A.B. da Silva 《Journal of The Franklin Institute》2018,355(4):1913-1930
Similarity-based modeling (SBM) is a technique whereby the normal operation of a system is modeled in order to detect faults by analyzing their similarity to the normal system states. First proposed around two decades ago, SBM has been successfully used for fault detection in varied systems. In spite of this success, there is not much study performed in the literature regarding its design, that encompasses both similarity metrics and model training. This work aims at contributing with an in-depth study of SBM for fault detection considering these two design aspects. This is done in the context of proposing a novel system to identify rotating-machinery faults based on SBM, that is employed either as a standalone classifier or to generate features for a random forest classifier. New approaches for training the model and new similarity metrics are investigated. Experimental results are shown for the recently developed Machinery Fault Database (MaFaulDa) that has an extensive set of sequences and fault types, and for the Case Western Reserve University (CWRU) bearing database. Results for both databases indicate that the proposed techniques increase the generalization power of the similarity model and of the associated classifier, achieving accuracies of 98.5% on MaFaulDa and 98.9% on CWRU database. 相似文献
14.
The classical probabilistic models attempt to capture the ad hoc information retrieval problem within a rigorous probabilistic framework. It has long been recognized that the primary obstacle to the effective performance of the probabilistic models is the need to estimate a relevance model. The Dirichlet compound multinomial (DCM) distribution based on the Polya Urn scheme, which can also be considered as a hierarchical Bayesian model, is a more appropriate generative model than the traditional multinomial distribution for text documents. We explore a new probabilistic model based on the DCM distribution, which enables efficient retrieval and accurate ranking. Because the DCM distribution captures the dependency of repetitive word occurrences, the new probabilistic model based on this distribution is able to model the concavity of the score function more effectively. To avoid the empirical tuning of retrieval parameters, we design several parameter estimation algorithms to automatically set model parameters. Additionally, we propose a pseudo-relevance feedback algorithm based on the mixture modeling of the Dirichlet compound multinomial distribution to further improve retrieval accuracy. Finally, our experiments show that both the baseline probabilistic retrieval algorithm based on the DCM distribution and the corresponding pseudo-relevance feedback algorithm outperform the existing language modeling systems on several TREC retrieval tasks. The main objective of this research is to develop an effective probabilistic model based on the DCM distribution. A secondary objective is to provide a thorough understanding of the probabilistic retrieval model by a theoretical understanding of various text distribution assumptions. 相似文献
15.
In the information management literature, Software-as-a-Service (SaaS) is recognized as a technology capable of providing operational and financial benefits to firms, and it is rising as the dominant IT service delivery model. Considered to be a promising solution it is garnering interest among researchers and professionals. However, SaaS can represent a vulnerability to firms due to its nature. The weighing of the pros and cons leads to firms’ uncertainty regarding SaaS adoption. Through the lenses of technology-organization-environment (TOE) framework we examine the contextual factors that influence the adoption of SaaS. Furthermore, this study explores the moderating effects of the environmental context in the adoption of SaaS and how it shapes the direct influences of technological and organizational contexts of the TOE framework. Data collected from 259 firms were used to test the proposed model. The study found the significance of the technology, organization, and environment context for SaaS adoption. Moreover, it was found the moderator influence of the environment context between the organization context and SaaS adoption. This study contributes to a deepest understanding of the determinants of SaaS adoption by providing a holistic theoretical lens, advancing newer paths of approaching the TOE framework. 相似文献
16.
《Information processing & management》2023,60(4):103348
Graph neural networks (GNN) have emerged as a new state-of-the-art for learning knowledge graph representations. Although they have shown impressive performance in recent studies, how to efficiently and effectively aggregate neighboring features is not well designed. To tackle this challenge, we propose the simplifying heterogeneous graph neural network (SHGNet), a generic framework that discards the two standard operations in GNN, including the transformation matrix and nonlinear activation. SHGNet, in particular, adopts only the essential component of neighborhood aggregation in GNN and incorporates relation features into feature propagation. Furthermore, to capture complex structures, SHGNet utilizes a hierarchical aggregation architecture, including node aggregation and relation weighting. Thus, the proposed model can treat each relation differently and selectively aggregate informative features. SHGNet has been evaluated for link prediction tasks on three real-world benchmark datasets. The experimental results show that SHGNet significantly promotes efficiency while maintaining superior performance, outperforming all the existing models in 3 out of 4 metrics on NELL-995 and in 4 out of 4 metrics on FB15k-237 dataset. 相似文献
17.
Developing a strong and positive information technology (IT) service climate is important to ensure high-quality IT services. The influence of job resources and personal resources on the IT service climate and the mediating effect of the IT service climate on IT service quality were explored in this study qualitatively and quantitatively. Among the various job resource factors, we found that training, reward, and information system (IS) support can improve the IT service climate. Coworker support shows a non-significant effect on IT service climate. Among the personal resource factors, self-efficacy and hope have significantly positive effects on the IT service climate. Optimism and resilience do not appear to significantly affect the IT service climate. Our results confirm that the IT service climate is an important mediator in enabling certain antecedents to properly impact IT service quality. The IT service climate was found to fully mediate the effects of training, reward, IS support, and hope while partially mediating the effect of self-efficacy. The effect of the IT service climate on IT service quality is also strengthened by a high customer contact frequency. The results presented here may provide scholars and practitioners with a comprehensive, workable understanding of the IT service climate. 相似文献
18.
《Information processing & management》2020,57(4):102248
In information retrieval, the task of query performance prediction (QPP) is concerned with determining in advance the performance of a given query within the context of a retrieval model. QPP has an important role in ensuring proper handling of queries with varying levels of difficulty. Based on the extant literature, query specificity is an important indicator of query performance and is typically estimated using corpus-specific frequency-based specificity metrics However, such metrics do not consider term semantics and inter-term associations. Our work presented in this paper distinguishes itself by proposing a host of corpus-independent specificity metrics that are based on pre-trained neural embeddings and leverage geometric relations between terms in the embedding space in order to capture the semantics of terms and their interdependencies. Specifically, we propose three classes of specificity metrics based on pre-trained neural embeddings: neighborhood-based, graph-based, and cluster-based metrics. Through two extensive and complementary sets of experiments, we show that the proposed specificity metrics (1) are suitable specificity indicators, based on the gold standards derived from knowledge hierarchies (Wikipedia category hierarchy and DMOZ taxonomy), and (2) have better or competitive performance compared to the state of the art QPP metrics, based on both TREC ad hoc collections namely Robust’04, Gov2 and ClueWeb’09 and ANTIQUE question answering collection. The proposed graph-based specificity metrics, especially those that capture a larger number of inter-term associations, proved to be the most effective in both query specificity estimation and QPP. We have also publicly released two test collections (i.e. specificity gold standards) that we built from the Wikipedia and DMOZ knowledge hierarchies. 相似文献
19.
20.
《Research Policy》2019,48(10):103557
Complex societal or environmental problems require fast and substantial socio-technical transitions. For instance, in the case of climate change, these transitions need to take place in the energy, transport and several industry sectors. To induce and accelerate such transitions, numerous policy interventions are required, which interact with each other in policy mixes. While several conceptual studies on policy mixes have been published recently, there is very little empirical research apart from single case or small-n studies. It has been prominently argued that the debate about policy mixes has reached an impasse partly due to this lack of empirical work. This paper addresses this gap by providing a first analysis of the temporal dynamics of complex policy mixes. To do so, we develop a conceptualization and measurement of policy mix balance across instrument types as well as policy mix design features (in the form of intensity as a general and technology specificity as a technology-focused design feature). This allows us to answer the question how temporal dynamics of policy mixes differ between countries regarding their balance and design features. Our measurement approach is developed bottom-up, i.e., policies are assessed individually and then aggregated systematically at the policy mix level. This enables overcoming the ‘dependent variable problem in the study of policy change’, i.e., the problem of measuring policy output. More specifically, we develop a comparative dataset of 522 renewable energy policies in nine OECD countries. Our analysis shows that countries’ policy mix dynamics vary strongly regarding some variables (e.g., technology specificity) but less regarding others (e.g., balance). As a validity check, we also test the effects of these mix dynamics on policy outcome in the form of renewable energy technology diffusion. We reflect our findings in light of the theoretical debates around policy mixes and policy design and discuss how our results provoke an agenda for the new generation of research on policy mixes. We specifically discuss avenues for future research with a particular focus on the ‘politics of policy mixes’. 相似文献