A Bayesian framework for learning shared and individual subspaces from multiple data sources

This paper presents a novel Bayesian formulation to exploit shared structures across multiple data sources, constructing foundations for effective mining and retrieval across disparate domains. We jointly analyze diverse data sources using a unifying piece of metadata (textual tags). We propose a method based on Bayesian Probabilistic Matrix Factorization (BPMF) which is able to explicitly model the partial knowledge common to the datasets using shared subspaces and the knowledge specific to each dataset using individual subspaces. For the proposed model, we derive an efficient algorithm for learning the joint factorization based on Gibbs sampling. The effectiveness of the model is demonstrated by social media retrieval tasks across single and multiple media. The proposed solution is applicable to a wider context, providing a formal framework suitable for exploiting individual as well as mutual knowledge present across heterogeneous data sources of many kinds.

Nature, extent and antecedents of risk managment practices in knowledge-intensive firms

This study provides empirical evidence on the nature and extent of risks faced by Small to Medium-Sized Knowledge Intensive Firms (SMKIFs) and the risk management approaches adopted by them. The study also assesses the effects of selected organisational factors such as industry, entity size and risk governance leadership on the commitment by SMKIFs to using an Enterprise Risk Management (ERM) approach. Data was obtained through a questionnaire survey of SMKIFs in the state of Victoria, Australia which were either in the bio-technology (bio-tech) or the accounting and legal (business services) industry sectors. Based on a total of 104 (13%) useable responses from senior managers in charge of risk management, some of the key findings include the identification of the top three risks faced by SMKIFs being (i) potential damage to firm’s reputation, (ii) inability to recruit and retain workers who have appropriate skills and expertise, and (iii) increase in costs. Interestingly, while 51% of the respondents described their firms as being willing to or keen to take risks, 38% saw their firms as being either preferring not to take risks or refuse to take risks, with the remainder of the firms (11%) viewed as neutral. The data also indicates that more than half of the respondent firms (54%) had established either a complete or a partial ERM system. Further, data analysis based on a binary logit regression model indicates bio-techs, firm size and directors’ support of risk management as key predictors of ERM implementation in SMKIFs.

Observation-switching linear dynamic systems for tracking humans through unexpected partial occlusions by scene objects

This paper focuses on the problem of tracking people through occlusions by scene objects. Rather than relying on models of the scene to predict when occlusions will occur as other researchers have done, this paper proposes a linear dynamic system that switches between two alternatives of the position measurement in order to handle occlusions as they occur. The filter automatically switches between a foot-based measure of position (assuming z = Q) to a head-based position measure (given the person's height) when an occlusion of the person's lower body occurs. No knowledge of the scene or its occluding objects is used. Unlike similar research [2, 14], the approach does not assume a fixed height for people and so is able to track humans through occlusions even when they change height during the occlusion. The approach is evaluated on three furnished scenes containing tables, chairs, desks and partitions. Occlusions range from occlusions of legs, occlusions whilst being seated and near-total occlusions where only the person's head is visible. Results show that the approach provides a significant reduction in false-positive tracks in a multi-camera environment, and more than halves the number of lost tracks in single monocular camera views.

Effects of entrepreneurial knowledge on entrepreneurial intentions: a longitudinal study of selected South-east Asian business students

Drawing on the theory of planned behaviour, this study examines the direct and indirect effects of knowledge gained from a formal entrepreneurship education programme on an individual’s entrepreneurial intentions (EI). It tracks the changes in students’ entrepreneurial knowledge (EK), perceptions of desirability of, and self-efficacy in, engaging in entrepreneurship and the impact of those changes on students’ EI upon completion of an entrepreneurship course. It uses longitudinal survey data of 245 business students in a Philippine university. Using cross-lagged panel method and partial-least squares-based structural equation modelling, the study builds and tests the measurement and structural models to examine the hypothesised interactions of EK, perceived desirability of, self-efficacy towards entrepreneurship, and EI. The findings underscore the importance of developing knowledge to nurture students’ self-confidence and attitudinal propensity to engage in entrepreneurship.

Effects of knowledge management on client-vendor relationship quality : the mediating role of global mindset

Purpose: This study argues that knowledge management (KM) by itself has only limited effects on client–vendor relationship (CVR) of global providers of highly customised services. Rather, it is the ability of top management to properly evaluate and utilise a vast array of complex knowledge which allows global firms to develop and maintain superior CVR. The paper tests the proposition that global mindset (GM) of top management mediates the effects of KM on CVR quality. Design/methodology/approach: The paper uses survey data from a sample of 68 international service providers (ISPs) in the information technology sector in India and partial least squares approach to structural equation modelling to test the hypotheses. Findings: The results show that both KM and GM have positive and statistically significant effects on the quality of CVRs. The results also confirm that the GM of top management has significant and substantive mediation effects on the relationship between KM and CVR quality. Research limitations/implications: The small size of the sample and the focus on ISPs in a single country constitute the main limitations of the study. Future research should ideally draw from a larger sample of ISPs from multiple countries and sectors in order to allow for greater generalisation of the findings. Practical implications: ISPs will benefit from developing the GM of their top management teams to enhance their CVRs. Originality/value: The paper provides new insights into how, in an international context, firms can transform their KM into superior CVR quality through the development of GM.