Search results

Educational data mining (EDM) discovers significant patterns from educational data and thus can help understand the relations between learners and their educational settings. However, most previous data mining techniques focus on prediction of learning performance of learners without integrating learning patterns identification techniques.

This study proposes a new framework for identifying learning patterns and predicting learning performance. Two modules, the learning patterns identification module and the deep learning prediction models (DNN), are integrated into this framework to identify the difference of learning performance and predicting learning performance from profiles of students.

Experimental results from survey data indicate that the proposed identifying learning patterns module could facilitate identifying valuable difference (change) patterns from student’s profiles. The proposed learning performance prediction module which adapts DNN also performs better than traditional machine techniques in prediction performance metrics.

To our best knowledge, the framework is the only educational system in the literature for identifying learning patterns and predicting learning performance.

On the one hand, this paper is to further understand the residents' differentiated power consumption behaviors and tap the residential family characteristics labels from the perspective of electricity stability. On the other hand, this paper is to address the problem of lack of causal relationship in the existing research on the association analysis of residential electricity consumption behavior and basic information data.

First, the density-based spatial clustering of applications with noise method is used to extract the typical daily load curve of residents. Second, the degree of electricity consumption stability is described from three perspectives: daily minimum load rate, daily load rate and daily load fluctuation rate, and is evaluated comprehensively using the entropy weight method. Finally, residential customer labels are constructed from sociological characteristics, residential characteristics and energy use attitudes, and the enhanced FP-growth algorithm is employed to investigate any potential links between each factor and the stability of electricity consumption.

Compared with the original FP-growth algorithm, the improved algorithm can realize the excavation of rules containing specific attribute labels, which improves the excavation efficiency. In terms of factors influencing electricity stability, characteristics such as a large number of family members, being well employed, having children in the household and newer dwelling labels may all lead to poorer electricity stability, but residents' attitudes toward energy use and dwelling type are not significantly associated with electricity stability.

This paper aims to uncover household socioeconomic traits that influence the stability of home electricity use and to shed light on the intricate connections between them. Firstly, in this article, from the perspective of electricity stability, the characteristics of the power consumption of residents' users are refined. And the authors use the entropy weight method to comprehensively evaluate the stability of electricity usage. Secondly, the labels of residential users' household characteristics are screened and organized. Finally, the improved FP-growth algorithm is used to mine the residential household characteristic labels that are strongly associated with electricity consumption stability.

1. The stability of electricity consumption is important to the stable operation of the grid.

2. An improved FP-growth algorithm is employed to explore the influencing factors.

3. The improved algorithm enables the mining of rules containing specific attribute labels.

4. Residents' attitudes toward energy use are largely unrelated to the stability of electricity use.

The stability of electricity consumption is important to the stable operation of the grid.

An improved FP-growth algorithm is employed to explore the influencing factors.

The improved algorithm enables the mining of rules containing specific attribute labels.

Residents' attitudes toward energy use are largely unrelated to the stability of electricity use.

This study aims to develop an interface management risk interaction modeling and analysis methodology applicable to complex systems in high-speed rail construction projects, reveal the interaction mechanism of interface management risk and provide theoretical support for project managers to develop appropriate interface management risk response strategies.

This paper introduces the association rule mining technique to improve the complex network modeling method. Taking China as an example, based on the stakeholder perspective, the risk factors and significant accident types of interface management of high-speed rail construction projects are systematically identified, and a database is established. Then, the Apriori algorithm is used to mine and analyze the strong association rules among the factors in the database, construct the complex network, and analyze its topological characteristics to reveal the interaction mechanism of the interface management risk of high-speed rail construction projects.

The results show that the network is both scale-free and small-world, implying that construction accidents are not random events but rather the result of strong interactions between numerous interface management risks. Contractors, technical interfaces, mechanical equipment, and environmental factors are the primary direct causal factors of accidents, while owners and designers are essential indirect causal factors. The global importance of stakeholders such as owners, designers, and supervisors rises significantly after considering the indirect correlations between factors. This theoretically explains the need to consider the interactions between interface management risks.

The interaction mechanism between interface management risks is unclear, which is an essential factor influencing the decision of risk response measures. This study proposes a new methodology for analyzing interface management risk response strategies that incorporate quantitative analysis methods and considers the interaction of interface management risks.

In order to estimate the value of semi-automated subject indexing in operative library catalogues, the study aimed to investigate five different automated implementations of an open source software package on a large set of Swedish union catalogue metadata records, with Dewey Decimal Classification (DDC) as the target classification system. It also aimed to contribute to the body of research on aboutness and related challenges in automated subject indexing and evaluation.

On a sample of over 230,000 records with close to 12,000 distinct DDC classes, an open source tool Annif, developed by the National Library of Finland, was applied in the following implementations: lexical algorithm, support vector classifier, fastText, Omikuji Bonsai and an ensemble approach combing the former four. A qualitative study involving two senior catalogue librarians and three students of library and information studies was also conducted to investigate the value and inter-rater agreement of automatically assigned classes, on a sample of 60 records.

The best results were achieved using the ensemble approach that achieved 66.82% accuracy on the three-digit DDC classification task. The qualitative study confirmed earlier studies reporting low inter-rater agreement but also pointed to the potential value of automatically assigned classes as additional access points in information retrieval.

The paper presents an extensive study of automated classification in an operative library catalogue, accompanied by a qualitative study of automated classes. It demonstrates the value of applying semi-automated indexing in operative information retrieval systems.

This paper aims to propose an incremental graph indexing method based on probabilistic inferences in Bayesian network (BN) for approximate nearest neighbor search (ANNS) that adds unindexed queries into the graph index incrementally.

This paper first uses the attention mechanism based graph convolutional network to embed a social network into the low-dimensional vector space, which could improve the efficiency of graph index construction. To add the unindexed queries into the graph index incrementally, this study proposes to learn the rule-based BN from social interactions. Thus, the dependency relations of unindexed queries and their neighbors are represented, and the probabilistic inferences in BN are then performed.

Experimental results demonstrate that the proposed method improves the search precision by at least 5% and search efficiency by 10% compared to the state-of-the-art methods.

This paper proposes a novel method to construct the incremental graph index based on probabilistic inferences in BN, such that both indexed and unindexed queries in ANNS could be addressed efficiently.

The identification of network user relationship in Fancircle contributes to quantifying the violence index of user text, mining the internal correlation of network behaviors among users, which provides necessary data support for the construction of knowledge graph.

A correlation identification method based on sentiment analysis (CRDM-SA) is put forward by extracting user semantic information, as well as introducing violent sentiment membership. To be specific, the topic of the implementation of topology mapping in the community can be obtained based on self-built field of violent sentiment dictionary (VSD) by extracting user text information. Afterward, the violence index of the user text is calculated to quantify the fuzzy sentiment representation between the user and the topic. Finally, the multi-granularity violence association rules mining of user text is realized by constructing violence fuzzy concept lattice.

It is helpful to reveal the internal relationship of online violence under complex network environment. In that case, the sentiment dependence of users can be characterized from a granular perspective.

The membership degree of violent sentiment into user relationship recognition in Fancircle community is introduced, and a text sentiment association recognition method based on VSD is proposed. By calculating the value of violent sentiment in the user text, the annotation of violent sentiment in the topic dimension of the text is achieved, and the partial order relation between fuzzy concepts of violence under the effective confidence threshold is utilized to obtain the association relation.

This study aims to characterize remote workers’ Airbnb experiences through the cognitive outcomes of their experiences and to consider the differences between long and short stays.

The structural topic model methodology was used to identify relevant topics. Data were collected from InsideAirbnb for Lisbon, Portugal and Austin, Texas, USA, for 2022 and early 2023, focusing on reviews that mentioned remote work.

The Airbnb experiences of remote workers and digital nomads are characterized as professionals who express mostly affective outcomes, but also have behavioral and nonaffective outcomes during their stay. In addition, the findings support the moderating role of length of stay and city.

This paper contributes to the literature by exploring how length of stay affects the priorities of remote workers on Airbnb, highlighting the different needs of long-term and short-term stays, and helping to consolidate and clarify the scattered research on customers’ long-term experiences in tourism and hospitality.

The Airbnb experience of remote workers is the highly valued as evidenced by the high rate of commending reviews indicating a willingness to stay there again. It is suggested that Airbnb hosts continue their helpful role and ensuring the functionality and availability of essential facilities and emphasizing neighborhood amenities specific to long and short stays. ChatGPT4 was found to be valuable for extracting data and assigning topic labels.

This study uses a novel structural topic model, augmented with ChatGPT4, to analyze Airbnb customer reviews that mention remote work, thereby improving inferences about the characterization of remote workers.

In the modern digital realm, while artificial intelligence (AI) technologies pave the way for unprecedented opportunities, they also give rise to intricate cybersecurity issues, including threats like deepfakes and unanticipated AI-induced risks. This study aims to address the insufficient exploration of AI cybersecurity awareness in the current literature.

Using in-depth surveys across varied sectors (N = 150), the authors analyzed the correlation between the absence of AI risk content in organizational cybersecurity awareness programs and its impact on employee awareness.

A significant AI-risk knowledge void was observed among users: despite frequent interaction with AI tools, a majority remain unaware of specialized AI threats. A pronounced knowledge difference existed between those that are trained in AI risks and those who are not, more apparent among non-technical personnel and sectors managing sensitive information.

This study paves the way for thorough research, allowing for refinement of awareness initiatives tailored to distinct industries.

It is imperative for organizations to emphasize AI risk training, especially among non-technical staff. Industries handling sensitive data should be at the forefront.

Ensuring employees are aware of AI-related threats can lead to a safer digital environment for both organizations and society at large, given the pervasive nature of AI in everyday life.

Unlike most of the papers about AI risks, the authors do not trust subjective data from second hand papers, but use objective authentic data from the authors’ own up-to-date anonymous survey.

This study surveyed the entrepreneurship community on a large university campus in 2016 and in 2020 to identify and understand its information habits and needs. User needs can inform service design and assessment, as well as inform approaches to reference interactions. Librarians are encouraged to conduct similar surveys to better understand this niche population.

Investigators employed repeated cross-sectional design, a longitudinal research approach that draws on samples of non-overlapping or minimally overlapping cases over time. Qualitative and quantitative data were collected using online survey instruments. Data collected included demographic information, venture characteristics, participation in institutional activities such as accelerator programs and credit courses, general startup research behaviors and needs, and details of a specific instance of business or market research as well as interaction with the library and access to training. Triangulation of semantic and episodic was applied to draw reliable conclusions about respondent behavior.

In both surveys, over half of respondents were students and 75% of respondents were engaged in startup activity, most at the early stages. While respondent demographics, type and purpose of information sought remained constant between the two surveys, awareness and use of the library rose on several metrics. Coding revealed insights into respondents’ attitudes toward and strategies for secondary business research. Information obtained during the research process had a moderate impact on their ventures. These findings informed the development of library research and instruction services, programs, and collections for entrepreneurs.

The repeated cross-sectional design of the study is unique and shows trends in the community over time. The mixed-methods approach provides a robust and nuanced portrait of the community. These findings informed the development and assessment of library research and instruction services, programs and collections for entrepreneurs.

This empirical study aims to explore how actors in specific human resource practices (HRPs) such as line managers (LMs) impact employee productivity measures in the context of financial institutions (FI) banks.

This cross-country study adopted a qualitative methodology. It employed semi-structured interviews to collect data from purposefully selected 12 business facing directors (BFDs) working in the top 10 banks in Nigeria and the UK. The data collected were analysed with the help of the trans-positional cognition approach (TPCA) phenomenological method.

The findings of a TPCA analytical process imply that in the UK and Nigeria’s FIs, the BFDs line managers’ human resources practices (LMHRPs) resulted in a highly regulated workplace, knowledge gap, service operations challenges and subjective quantitatively driven key performance indicators, considered service productivity paradoxical elements. Although the practices in the UK and Nigerian FIs had similar labels, their aggregates were underpinned by different contextual issues.

To support LMs in better understanding and managing FIs BFDs productivity measures and outcomes, we propose the Managerial Employee Productivity Operational Definition framework as part of their toolkit. This study will be helpful for banking sectors, their regulators, policymakers, other FIs’ industry stakeholders and future researchers in the field.

Within the context of the UK and Nigeria’s FIs, this study is the first attempt to understand how LMHRPs impact BFDs productivity in this manner. It confirms that LMHRPs result in service productivity paradoxical elements with perceived or lost productivity implications.