| 1. |
- Bergström, Erik, 1976-, et al.
(författare)
-
Developing an information classification method
- 2021
-
Ingår i: Information and Computer Security. - : Emerald Group Publishing Limited. - 2056-4961. ; 29:2, s. 209-239
-
Tidskriftsartikel (refereegranskat)abstract
- Purpose: The purpose of this paper is to develop a method for information classification. The proposed method draws on established standards, such as the ISO/IEC 27002 and information classification practices. The long-term goal of the method is to decrease the subjective judgement in the implementation of information classification in organisations, which can lead to information security breaches because the information is under- or over-classified. Design/methodology/approach: The results are based on a design science research approach, implemented as five iterations spanning the years 2013 to 2019. Findings: The paper presents a method for information classification and the design principles underpinning the method. The empirical demonstration shows that senior and novice information security managers perceive the method as a useful tool for classifying information assets in an organisation. Research limitations/implications: Existing research has, to a limited extent, provided extensive advice on how to approach information classification in organisations systematically. The method presented in this paper can act as a starting point for further research in this area, aiming at decreasing subjectivity in the information classification process. Additional research is needed to fully validate the proposed method for information classification and its potential to reduce the subjective judgement. Practical implications: The research contributes to practice by offering a method for information classification. It provides a hands-on-tool for how to implement an information classification process. Besides, this research proves that it is possible to devise a method to support information classification. This is important, because, even if an organisation chooses not to adopt the proposed method, the very fact that this method has proved useful should encourage any similar endeavour. Originality/value: The proposed method offers a detailed and well-elaborated tool for information classification. The method is generic and adaptable, depending on organisational needs.
|
|
| 2. |
- Koriakina, Nadezhda, 1991-, et al.
(författare)
-
Deep multiple instance learning versus conventional deep single instance learning for interpretable oral cancer detection
- 2024
-
Ingår i: PLOS ONE. - : Public Library of Science (PLoS). - 1932-6203. ; 19:4 April
-
Tidskriftsartikel (refereegranskat)abstract
- The current medical standard for setting an oral cancer (OC) diagnosis is histological examination of a tissue sample taken from the oral cavity. This process is time-consuming and more invasive than an alternative approach of acquiring a brush sample followed by cytological analysis. Using a microscope, skilled cytotechnologists are able to detect changes due to malignancy; however, introducing this approach into clinical routine is associated with challenges such as a lack of resources and experts. To design a trustworthy OC detection system that can assist cytotechnologists, we are interested in deep learning based methods that can reliably detect cancer, given only per-patient labels (thereby minimizing annotation bias), and also provide information regarding which cells are most relevant for the diagnosis (thereby enabling supervision and understanding). In this study, we perform a comparison of two approaches suitable for OC detection and interpretation: (i) conventional single instance learning (SIL) approach and (ii) a modern multiple instance learning (MIL) method. To facilitate systematic evaluation of the considered approaches, we, in addition to a real OC dataset with patient-level ground truth annotations, also introduce a synthetic dataset—PAP-QMNIST. This dataset shares several properties of OC data, such as image size and large and varied number of instances per bag, and may therefore act as a proxy model of a real OC dataset, while, in contrast to OC data, it offers reliable per-instance ground truth, as defined by design. PAP-QMNIST has the additional advantage of being visually interpretable for non-experts, which simplifies analysis of the behavior of methods. For both OC and PAP-QMNIST data, we evaluate performance of the methods utilizing three different neural network architectures. Our study indicates, somewhat surprisingly, that on both synthetic and real data, the performance of the SIL approach is better or equal to the performance of the MIL approach. Visual examination by cytotechnologist indicates that the methods manage to identify cells which deviate from normality, including malignant cells as well as those suspicious for dysplasia. We share the code as open source.
|
|
| 3. |
- Sweidan, Dirar, et al.
(författare)
-
Predicting Customer Churn in Retailing
- 2022
-
Ingår i: Proceedings 21st IEEE International Conference on Machine Learning and Applications ICMLA 2022. - : IEEE. - 9781665462839 - 9781665462846 ; , s. 635-640
-
Konferensbidrag (refereegranskat)abstract
- Customer churn is one of the most challenging problems for digital retailers. With significantly higher costs for acquiring new customers than retaining existing ones, knowledge about which customers are likely to churn becomes essential. This paper reports a case study where a data-driven approach to churn prediction is used for predicting churners and gaining insights about the problem domain. The real-world data set used contains approximately 200 000 customers, describing each customer using more than 50 features. In the pre-processing, exploration, modeling and analysis, attributes related to recency, frequency, and monetary concepts are identified and utilized. In addition, correlations and feature importance are used to discover and understand churn indicators. One important finding is that the churn rate highly depends on the number of previous purchases. In the segment consisting of customers with only one previous purchase, more than 75% will churn, i.e., not make another purchase in the coming year. For customers with at least four previous purchases, the corresponding churn rate is around 25%. Further analysis shows that churning customers in general, and as expected, make smaller purchases and visit the online store less often. In the experimentation, three modeling techniques are evaluated, and the results show that, in particular, Gradient Boosting models can predict churners with relatively high accuracy while obtaining a good balance between precision and recall.
|
|
| 4. |
- Beckerman, Carina, 1956-
(författare)
-
Implications of Transforming the Patient Record into a Knowledge Management System : Initiating a Movement of Coordination and Enhancement
- 2008
-
Ingår i: The ICFAI University Journal of Knowledge Management. - New Dehli : The ICFAI University Press. - 0972-9216. ; Nov:6
-
Tidskriftsartikel (refereegranskat)abstract
- Today there is often a need to re-innovate who you are and what you do and re-think the tools that are used and the business models that guide action. The purpose of this paper is to show how transforming a document, such as a patient record, might start a horizontal and vertical movement, a movement of coordination and enhancement in an organizational setting, such as a hospital clinic. The observations presented here and the conclusions drawn were obtained during a three year case study following implications of constructing and computerizing a patient record at three different hospitals. The results were then analyzed, interpreted and discussed within a framework combining theories about knowledge management and with cognitive theories about use of interpretative schemes and representations. This paper tries to look beyond the implications of reconstructing a patient record on a micro-level or explore if it is good or bad to computerize it. Instead this paper theorizes about how re-thinking the interpretative scheme for what a patient record is and how it may be used might restructure a health care setting. It proposes that what the employees want to achieve with the knowledge management system depends on what strategy they have for it.
|
|
| 5. |
- Kävrestad, Joakim, 1989-, et al.
(författare)
-
The impact of short-term memory on phishing detection ability and password behaviour
- 2023
-
Ingår i: Proceedings of the 9th International Conference on Socio-Technical Perspective in Information Systems Development (STPIS 2023). - : CEUR-WS. ; , s. 160-173, s. 160-173
-
Konferensbidrag (refereegranskat)abstract
- Cybersecurity is a socio-technical discipline which is dependent on the interplay between users and devices, and the organizations where this interplay takes place. Previous research has shown that the interplay between users and devices is highly affected by the cognitive abilities of users. This is prominent in cybersecurity, which requires users to make security-aware decisions when, for instance, reading emails and decide which emails are legitimate and which emails constitute phishing. Research further suggests that decision-making is dependent on memory ability, which is the focus of this research. In this study, we investigate the impact of short-term memory on phishing detection ability and password behaviour. A web survey was used to collect quantitative data from a large sample of respondents. The survey was distributed on social media platforms and 93 participants completed the survey. The results indicate a positive correlation between short-term memory scores and both password detection ability and password behavior.
|
|
| 6. |
- Salomonson, Nicklas, et al.
(författare)
-
Comparing Human-to-Human and Human-to-AEA Communication in Service Encounters
- 2013
-
Ingår i: Journal of Business Communication. - : SAGE. - 0021-9436 .- 1552-4582. ; 50:1, s. 87-116
-
Tidskriftsartikel (refereegranskat)abstract
- An increasing number of companies are introducing artificial agents as self-service tools on their websites, often motivated by the need to provide cost-efficient interaction solutions. These agents are designed to help customers and clients to conduct their business on the website. Their role on commercial websites is often to act as online sales/shopping assistants with the hope of replacing some of the interactions between customers and sales staff, thus supplementing or replacing human-to-human communication. However, research on artificial agents and comparisons with human-to-human communication, in particular, is still scarce. The purpose of this article is to explore the similarities and differences in communication between an artificial agent and customers compared with face-to-face communication between human service providers and customers. The method employed is a qualitative comparison of face-to-face human service provision in a travel agency setting and logs of interactions between customers and an artificial agent on an airline company website. The analysis is based on the theory of “activity-based communication analysis” and makes use of a framework of specific communication features provided by this theory. The article demonstrates a number of deficiencies in communication between artificial embodied agents and humans, suggesting that artificial embodied agents still lack many of the desirable communicative aspects of human-to-human service encounters.
|
|
| 7. |
- Johansson, Ulf, et al.
(författare)
-
Overproduce-and-Select : The Grim Reality
- 2013
-
Ingår i: 2013 IEEE Symposium on Computational Intelligence and Ensemble Learning (CIEL). - : IEEE. - 9781467358538 ; , s. 52-59
-
Konferensbidrag (refereegranskat)abstract
- Overproduce-and-select (OPAS) is a frequently used paradigm for building ensembles. In static OPAS, a large number of base classifiers are trained, before a subset of the available models is selected to be combined into the final ensemble. In general, the selected classifiers are supposed to be accurate and diverse for the OPAS strategy to result in highly accurate ensembles, but exactly how this is enforced in the selection process is not obvious. Most often, either individual models or ensembles are evaluated, using some performance metric, on available and labeled data. Naturally, the underlying assumption is that an observed advantage for the models (or the resulting ensemble) will carry over to test data. In the experimental study, a typical static OPAS scenario, using a pool of artificial neural networks and a number of very natural and frequently used performance measures, is evaluated on 22 publicly available data sets. The discouraging result is that although a fairly large proportion of the ensembles obtained higher test set accuracies, compared to using the entire pool as the ensemble, none of the selection criteria could be used to identify these highly accurate ensembles. Despite only investigating a specific scenario, we argue that the settings used are typical for static OPAS, thus making the results general enough to question the entire paradigm.
|
|
| 8. |
- Linnusson, Henrik, et al.
(författare)
-
Efficient conformal predictor ensembles
- 2020
-
Ingår i: Neurocomputing. - : Elsevier BV. - 0925-2312 .- 1872-8286. ; 397, s. 266-278
-
Tidskriftsartikel (refereegranskat)abstract
- In this paper, we study a generalization of a recently developed strategy for generating conformal predictor ensembles: out-of-bag calibration. The ensemble strategy is evaluated, both theoretically and empirically, against a commonly used alternative ensemble strategy, bootstrap conformal prediction, as well as common non-ensemble strategies. A thorough analysis is provided of out-of-bag calibration, with respect to theoretical validity, empirical validity (error rate), efficiency (prediction region size) and p-value stability (the degree of variance observed over multiple predictions for the same object). Empirical results show that out-of-bag calibration displays favorable characteristics with regard to these criteria, and we propose that out-of-bag calibration be adopted as a standard method for constructing conformal predictor ensembles.
|
|
| 9. |
- Sandkuhl, Kurt, 1963-, et al.
(författare)
-
Supporting early phases of digital twin development with enterprise modeling and capability management : Requirements from two industrial cases
- 2020
-
Ingår i: Enterprise, Business-Process and Information Systems Modeling. - Cham : Springer. - 9783030494179 - 9783030494186 ; , s. 284-299
-
Konferensbidrag (refereegranskat)abstract
- Industry 4.0 is a concept that has attracted much research and development over the last decade. At its core is the need to connect physical devices with their digital representations which essentially means establishing a digital twin. Currently, the technological development of digital twins has gathered much attention while the organizational and business aspects are less investigated. In response, the suitability of enterprise modeling and capability management for the purpose of developing and management of business-driven digital twins has been analyzed. A number of requirements from literature are summarized and two industrial cases have been analyzed for the purpose of investigating how the digital twin initiatives emerge and what forces drive the start of their implementation projects. The findings are discussed with respect to how Enterprise Modeling and the Capability-Driven Development method are able to support the business motivation, design and runtime management of digital twins.
|
|
| 10. |
- Huhnstock, Nikolas Alexander, 1988-, et al.
(författare)
-
An Infinite Replicated Softmax Model for Topic Modeling
- 2019
-
Ingår i: Modeling Decisions for Artificial Intelligence. - Cham : Springer. - 9783030267728 - 9783030267735 ; , s. 307-318
-
Konferensbidrag (refereegranskat)abstract
- In this paper, we describe the infinite replicated Softmax model (iRSM) as an adaptive topic model, utilizing the combination of the infinite restricted Boltzmann machine (iRBM) and the replicated Softmax model (RSM). In our approach, the iRBM extends the RBM by enabling its hidden layer to adapt to the data at hand, while the RSM allows for modeling low-dimensional latent semantic representation from a corpus. The combination of the two results is a method that is able to self-adapt to the number of topics within the document corpus and hence, renders manual identification of the correct number of topics superfluous. We propose a hybrid training approach to effectively improve the performance of the iRSM. An empirical evaluation is performed on a standard data set and the results are compared to the results of a baseline topic model. The results show that the iRSM adapts its hidden layer size to the data and when trained in the proposed hybrid manner outperforms the base RSM model.
|
|
