Change search
Link to record
Permanent link

Direct link
BETA
Alternative names
Publications (10 of 42) Show all publications
Boldt, M. & Rekanar, K. (2019). Analysis and text classification of privacy policies from rogue and top-100 fortune global companies. International Journal of Information Security and Privacy, 13(2), 47-66
Open this publication in new window or tab >>Analysis and text classification of privacy policies from rogue and top-100 fortune global companies
2019 (English)In: International Journal of Information Security and Privacy, ISSN 1930-1650, E-ISSN 1930-1669, Vol. 13, no 2, p. 47-66Article in journal (Refereed) Published
Abstract [en]

In the present article, the authors investigate to what extent supervised binary classification can be used to distinguish between legitimate and rogue privacy policies posted on web pages. 15 classification algorithms are evaluated using a data set that consists of 100 privacy policies from legitimate websites (belonging to companies that top the Fortune Global 500 list) as well as 67 policies from rogue websites. A manual analysis of all policy content was performed and clear statistical differences in terms of both length and adherence to seven general privacy principles are found. Privacy policies from legitimate companies have a 98% adherence to the seven privacy principles, which is significantly higher than the 45% associated with rogue companies. Out of the 15 evaluated classification algorithms, Naïve Bayes Multinomial is the most suitable candidate to solve the problem at hand. Its models show the best performance, with an AUC measure of 0.90 (0.08), which outperforms most of the other candidates in the statistical tests used. Copyright © 2019, IGI Global.

Place, publisher, year, edition, pages
IGI Global, 2019
Keywords
Classification, Classification algorithms, Information security, Machine learning, Privacy policies, Privacy policy data set, Data mining, Data privacy, Learning systems, Security of data, Text processing, Websites, Binary classification, Classification algorithm, Data set, Fortune global 500, Privacy principle, Statistical differences, Text classification, Classification (of information)
National Category
Computer Sciences
Identifiers
urn:nbn:se:bth-17875 (URN)10.4018/IJISP.2019040104 (DOI)000467764600004 ()2-s2.0-85064536690 (Scopus ID)
Available from: 2019-05-02 Created: 2019-05-02 Last updated: 2019-06-13Bibliographically approved
Boldt, M., Borg, A., Ickin, S. & Gustafsson, J. (2019). Anomaly detection of event sequences using multiple temporal resolutions and Markov chains. Knowledge and Information Systems
Open this publication in new window or tab >>Anomaly detection of event sequences using multiple temporal resolutions and Markov chains
2019 (English)In: Knowledge and Information Systems, ISSN 0219-1377, E-ISSN 0219-3116Article in journal (Refereed) Epub ahead of print
Abstract [en]

Streaming data services, such as video-on-demand, are getting increasingly more popular, and they are expected to account for more than 80% of all Internet traffic in 2020. In this context, it is important for streaming service providers to detect deviations in service requests due to issues or changing end-user behaviors in order to ensure that end-users experience high quality in the provided service. Therefore, in this study we investigate to what extent sequence-based Markov models can be used for anomaly detection by means of the end-users’ control sequences in the video streams, i.e., event sequences such as play, pause, resume and stop. This anomaly detection approach is further investigated over three different temporal resolutions in the data, more specifically: 1 h, 1 day and 3 days. The proposed anomaly detection approach supports anomaly detection in ongoing streaming sessions as it recalculates the probability for a specific session to be anomalous for each new streaming control event that is received. Two experiments are used for measuring the potential of the approach, which gives promising results in terms of precision, recall, F 1 -score and Jaccard index when compared to k-means clustering of the sessions. © 2019, The Author(s).

Place, publisher, year, edition, pages
Springer London, 2019
Keywords
Anomaly detection, Event sequences, Markov Chains, Multiple temporal resolutions, Video-on-demand
National Category
Computer Sciences
Identifiers
urn:nbn:se:bth-18026 (URN)10.1007/s10115-019-01365-y (DOI)2-s2.0-85066031197 (Scopus ID)
Available from: 2019-06-14 Created: 2019-06-14 Last updated: 2019-06-17Bibliographically approved
Borg, A., Boldt, M. & Svensson, J. (2019). Using conformal prediction for multi-label document classification in e-Mail support systems. In: Wotawa, F; Friedrich, G; Pill, I; KoitzHristov, R; Ali, M (Ed.), ADVANCES AND TRENDS IN ARTIFICIAL INTELLIGENCE: FROM THEORY TO PRACTICE: . Paper presented at International Conference on Computational Science, ICCS, Faro, Algarve, 9 July 2019 through 11 July 2019 (pp. 308-322). Springer Verlag, 11536
Open this publication in new window or tab >>Using conformal prediction for multi-label document classification in e-Mail support systems
2019 (English)In: ADVANCES AND TRENDS IN ARTIFICIAL INTELLIGENCE: FROM THEORY TO PRACTICE / [ed] Wotawa, F; Friedrich, G; Pill, I; KoitzHristov, R; Ali, M, Springer Verlag , 2019, Vol. 11536, p. 308-322Conference paper, Published paper (Refereed)
Abstract [en]

For any corporation the interaction with its customers is an important business process. This is especially the case for resolving various business-related issues that customers encounter. Classifying the type of such customer service e-mails to provide improved customer service is thus important. The classification of e-mails makes it possible to direct them to the most suitable handler within customer service. We have investigated the following two aspects of customer e-mail classification within a large Swedish corporation. First, whether a multi-label classifier can be introduced that performs similarly to an already existing multi-class classifier. Second, whether conformal prediction can be used to quantify the certainty of the predictions without loss in classification performance. Experiments were used to investigate these aspects using several evaluation metrics. The results show that for most evaluation metrics, there is no significant difference between multi-class and multi-label classifiers, except for Hamming loss where the multi-label approach performed with a lower loss. Further, the use of conformal prediction did not introduce any significant difference in classification performance for neither the multi-class nor the multi-label approach. As such, the results indicate that conformal prediction is a useful addition that quantifies the certainty of predictions without negative effects on the classification performance, which in turn allows detection of statistically significant predictions. © Springer Nature Switzerland AG 2019.

Place, publisher, year, edition, pages
Springer Verlag, 2019
Series
Lecture Notes in Computer Science, ISSN 0302-9743, E-ISSN 1611-3349 ; 11536
Keywords
Conformal prediction, Customer support e-mail, Multi-label classification, Electronic mail, Forecasting, Information retrieval systems, Intelligent systems, Sales, Classification performance, Conformal predictions, Customer support, Document Classification, Email classification, Evaluation metrics, Multi label classification, Multi-class classifier, Classification (of information)
National Category
Computer Sciences
Identifiers
urn:nbn:se:bth-18592 (URN)10.1007/978-3-030-22999-3_28 (DOI)000494939200028 ()2-s2.0-85068624865 (Scopus ID)9783030229986 (ISBN)
Conference
International Conference on Computational Science, ICCS, Faro, Algarve, 9 July 2019 through 11 July 2019
Available from: 2019-09-09 Created: 2019-09-09 Last updated: 2019-12-13Bibliographically approved
Boldt, M., Boeva, V. & Borg, A. (2018). Multi-expert estimations of burglars' risk exposure and level of pre-crime preparation using coded crime scene data: Work in progress. In: Brynielsson, J (Ed.), Proceedings - 2018 European Intelligence and Security Informatics Conference, EISIC 2018: . Paper presented at 8th European Intelligence and Security Informatics Conference, EISIC, Karlskrona, 24 October 2018 through 25 October 2018 (pp. 77-80). Institute of Electrical and Electronics Engineers Inc.
Open this publication in new window or tab >>Multi-expert estimations of burglars' risk exposure and level of pre-crime preparation using coded crime scene data: Work in progress
2018 (English)In: Proceedings - 2018 European Intelligence and Security Informatics Conference, EISIC 2018 / [ed] Brynielsson, J, Institute of Electrical and Electronics Engineers Inc. , 2018, p. 77-80Conference paper, Published paper (Refereed)
Abstract [en]

Law enforcement agencies strive to link crimes perpetrated by the same offenders into crime series in order to improve investigation efficiency. Such crime linkage can be done using both physical traces (e.g., DNA or fingerprints) or 'soft evidence' in the form of offenders' modus operandi (MO), i.e. their behaviors during crimes. However, physical traces are only present for a fraction of crimes, unlike behavioral evidence. This work-in-progress paper presents a method for aggregating multiple criminal profilers' ratings of offenders' behavioral characteristics based on feature-rich crime scene descriptions. The method calculates consensus ratings from individual experts' ratings, which then are used as a basis for classification algorithms. The classification algorithms can automatically generalize offenders' behavioral characteristics from cues in the crime scene data. Models trained on the consensus rating are evaluated against models trained on individual profiler's ratings. Thus, whether the consensus model shows improved performance over individual models. © 2018 IEEE.

Place, publisher, year, edition, pages
Institute of Electrical and Electronics Engineers Inc., 2018
Keywords
classification, crime linkage, Multi-expert decision making, offender profiling, Classification (of information), Risk management, Risk perception, Behavioral characteristics, Classification algorithm, Consensus models, Individual models, Law-enforcement agencies, Work in progress, Crime
National Category
Computer Sciences
Identifiers
urn:nbn:se:bth-18621 (URN)10.1109/EISIC.2018.00021 (DOI)000483031300012 ()2-s2.0-85069498311 (Scopus ID)9781538694008 (ISBN)
Conference
8th European Intelligence and Security Informatics Conference, EISIC, Karlskrona, 24 October 2018 through 25 October 2018
Available from: 2019-09-10 Created: 2019-09-10 Last updated: 2019-09-19Bibliographically approved
Boldt, M., Borg, A., Svensson, M. & Hildeby, J. (2018). Predicting burglars' risk exposure and level of pre-crime preparation using crime scene data. Intelligent Data Analysis, 22(1), 167-190, Article ID IDA 322-3210.
Open this publication in new window or tab >>Predicting burglars' risk exposure and level of pre-crime preparation using crime scene data
2018 (English)In: Intelligent Data Analysis, ISSN 1088-467X, Vol. 22, no 1, p. 167-190, article id IDA 322-3210Article in journal (Refereed) Published
Abstract [en]

Objectives: The present study aims to extend current research on how offenders’ modus operandi (MO) can be used in crime linkage, by investigating the possibility to automatically estimate offenders’ risk exposure and level of pre-crime preparation for residential burglaries. Such estimations can assist law enforcement agencies when linking crimes into series and thus provide a more comprehensive understanding of offenders and targets, based on the combined knowledge and evidence collected from different crime scenes. Methods: Two criminal profilers manually rated offenders’ risk exposure and level of pre-crime preparation for 50 burglaries each. In an experiment we then analyzed to what extent 16 machine-learning algorithms could generalize both offenders’ risk exposure and preparation scores from the criminal profilers’ ratings onto 15,598 residential burglaries. All included burglaries contain structured and feature-rich crime descriptions which learning algorithms can use to generalize offenders’ risk and preparation scores from.Results: Two models created by Naïve Bayes-based algorithms showed best performance with an AUC of 0.79 and 0.77 for estimating offenders' risk and preparation scores respectively. These algorithms were significantly better than most, but not all, algorithms. Both scores showed promising distinctiveness between linked series, as well as consistency for crimes within series compared to randomly sampled crimes.Conclusions: Estimating offenders' risk exposure and pre-crime preparation  can complement traditional MO characteristics in the crime linkage process. The estimations are also indicative to function for cross-category crimes that otherwise lack comparable MO. Future work could focus on increasing the number of manually rated offenses as well as fine-tuning the Naïve Bayes algorithm to increase its estimation performance.

Place, publisher, year, edition, pages
IOS Press, 2018
Keywords
Predictive models, Classification, Crime linkage, Offender behavior, Serial crime, Residential burglary
National Category
Computer Sciences
Identifiers
urn:nbn:se:bth-13935 (URN)10.3233/IDA-163220 (DOI)000426790500009 ()
Available from: 2017-02-21 Created: 2017-02-21 Last updated: 2018-04-04Bibliographically approved
Borg, A., Boldt, M. & Eliasson, J. (2017). Detecting Crime Series Based on Route Estimation and Behavioral Similarity. In: Brynielsson, J (Ed.), 2017 EUROPEAN INTELLIGENCE AND SECURITY INFORMATICS CONFERENCE (EISIC): . Paper presented at European Intelligence and Security Informatics Conference (EISIC), Athens (pp. 1-8). IEEE
Open this publication in new window or tab >>Detecting Crime Series Based on Route Estimation and Behavioral Similarity
2017 (English)In: 2017 EUROPEAN INTELLIGENCE AND SECURITY INFORMATICS CONFERENCE (EISIC) / [ed] Brynielsson, J, IEEE , 2017, p. 1-8Conference paper, Published paper (Refereed)
Abstract [en]

A majority of crimes are committed by a minority of offenders. Previous research has provided some support for the theory that serial offenders leave behavioral traces on the crime scene which could be used to link crimes to serial offenders. The aim of this work is to investigate to what extent it is possible to use geographic route estimations and behavioral data to detect serial offenders. Experiments were conducted using behavioral data from authentic burglary reports to investigate if it was possible to find crime routes with high similarity. Further, the use of burglary reports from serial offenders to investigate to what extent it was possible to detect serial offender crime routes. The result show that crime series with the same offender on average had a higher behavioral similarity than random crime series. Sets of crimes with high similarity, but without a known offender would be interesting for law enforcement to investigate further. The algorithm is also evaluated on 9 crime series containing a maximum of 20 crimes per series. The results suggest that it is possible to detect crime series with high similarity using analysis of both geographic routes and behavioral data recorded at crime scenes.

Place, publisher, year, edition, pages
IEEE, 2017
Series
European Intelligence and Security Informatics Conference, ISSN 2572-3723
Keywords
Crime route analysis, crime linkage, residential burglary, Behavioral analysis
National Category
Computer Sciences
Identifiers
urn:nbn:se:bth-15985 (URN)10.1109/EISIC.2017.10 (DOI)000425928200001 ()978-1-5386-2385-5 (ISBN)
Conference
European Intelligence and Security Informatics Conference (EISIC), Athens
Available from: 2018-03-23 Created: 2018-03-23 Last updated: 2018-05-18Bibliographically approved
Borg, A., Boldt, M. & Eliasson, J. (2017). Detecting Crime Series Based on Route Estimationand Behavioral Similarity. In: : . Paper presented at European Intelligence and Security Informatics Conference (EISIC), Attica, Greece. IEEE Computer Society
Open this publication in new window or tab >>Detecting Crime Series Based on Route Estimationand Behavioral Similarity
2017 (English)Conference paper, Published paper (Refereed)
Abstract [en]

A majority of crimes are committed by a minority of offenders. Previous research has provided some support for the theory that serial offenders leave behavioral traces on the crime scene which could be used to link crimes to serial offenders. The aim of this work is to investigate to what extent it is possible to use geographic route estimations and behavioral data to detect serial offenders. Experiments were conducted using behavioral data from authentic burglary reports to investigate if it was possible to find crime routes with high similarity. Further, the use of burglary reports from serial offenders to investigate to what extent it was possible to detect serial offender crime routes. The result show that crime series with the same offender on average had a higher behavioral similarity than random crime series. Sets of crimes with high similarity, but without a known offender would be interesting for law enforcement to investigate further. The algorithm is also evaluated on 9 crime series containing a maximum of 20 crimes per series. The results suggest that it is possible to detect crime series with high similarity using analysis of both geographic routes and behavioral data recorded at crime scenes.

Place, publisher, year, edition, pages
IEEE Computer Society, 2017
Keywords
Crime route analysis, crime linkage, residential burglary, Behavioral analysis
National Category
Computer Sciences
Identifiers
urn:nbn:se:bth-15167 (URN)
Conference
European Intelligence and Security Informatics Conference (EISIC), Attica, Greece
Available from: 2017-09-20 Created: 2017-09-20 Last updated: 2018-01-13Bibliographically approved
Erlandsson, F., Bródka, P., Boldt, M. & Johnson, H. (2017). Do We Really Need To Catch Them All?: A New User-Guided Social Media Crawling Method. Entropy, 19(12), Article ID 686.
Open this publication in new window or tab >>Do We Really Need To Catch Them All?: A New User-Guided Social Media Crawling Method
2017 (English)In: Entropy, ISSN 1099-4300, E-ISSN 1099-4300, Vol. 19, no 12, article id 686Article in journal (Refereed) Published
Abstract [en]

With the growing use of popular social media services like Facebook and Twitter it is hard to collect all content from the networks without access to the core infrastructure or paying for it. Thus, if all content cannot be collected one must consider which data are of most importance.In this work we present a novel User-Guided Social Media Crawling method (USMC) that is able to collect data from social media, utilizing the wisdom of the crowd to decide the order in which user generated content should be collected, to cover as many user interactions as possible. USMC is validated by crawling 160 Facebook public pages, containing 368 million users and 1.3 billion interactions, and it is compared with two other crawling methods. The results show that it is possible to cover approximately 75% of the interactions on a Facebook page by sampling just 20% of its posts, and at the same time reduce the crawling time by 53%.What is more, the social network constructed from the 20% sample has more than 75% of the users and edges compared to the social network created from all posts, and has very similar degree distribution.

Place, publisher, year, edition, pages
MDPI AG, 2017
Keywords
social media, social networks, sampling, crawling
National Category
Computer Sciences
Identifiers
urn:nbn:se:bth-15508 (URN)10.3390/e19120686 (DOI)000419007900055 ()
Available from: 2017-11-15 Created: 2017-11-15 Last updated: 2018-01-26Bibliographically approved
Jacobsson, A., Boldt, M. & Carlsson, B. (2016). A risk analysis of a smart home automation system. Future generations computer systems, 56, 719-733
Open this publication in new window or tab >>A risk analysis of a smart home automation system
2016 (English)In: Future generations computer systems, ISSN 0167-739X, E-ISSN 1872-7115, Vol. 56, p. 719-733Article in journal (Refereed) Published
Abstract [en]

Enforcing security in Internet of Things environments has been identified as one of the top barriers for realizing the vision of smart, energy-efficient homes and buildings. In this context, understanding the risks related to the use and potential misuse of information about homes, partners, and end-users, as well as, forming methods for integrating security-enhancing measures in the design is not straightforward and thus requires substantial investigation. A risk analysis applied on a smart home automation system developed in a research project involving leading industrial actors has been conducted. Out of 32 examined risks, 9 were classified as low and 4 as high, i.e., most of the identified risks were deemed as moderate. The risks classified as high were either related to the human factor or to the software components of the system. The results indicate that with the implementation of standard security features, new, as well as, current risks can be minimized to acceptable levels albeit that the most serious risks, i.e., those derived from the human factor, need more careful consideration, as they are inherently complex to handle. A discussion of the implications of the risk analysis results points to the need for a more general model of security and privacy included in the design phase of smart homes. With such a model of security and privacy in design in place, it will contribute to enforcing system security and enhancing user privacy in smart homes, and thus helping to further realize the potential in such loT environments. (C) 2015 Elsevier B.V. All rights reserved.

Place, publisher, year, edition, pages
Elsevier, 2016
Keywords
Internet of Things, Smart home automation, Risk analys, Privacy, Security
National Category
Computer Sciences
Identifiers
urn:nbn:se:bth-11661 (URN)10.1016/j.future.2015.09.003 (DOI)000368652500060 ()
Available from: 2016-03-02 Created: 2016-02-29 Last updated: 2018-01-10Bibliographically approved
Borg, A. & Boldt, M. (2016). Clustering Residential Burglaries Using Modus Operandi and Spatiotemporal Information. International Journal of Information Technology and Decision Making, 15(1), 23-42
Open this publication in new window or tab >>Clustering Residential Burglaries Using Modus Operandi and Spatiotemporal Information
2016 (English)In: International Journal of Information Technology and Decision Making, ISSN 0219-6220, Vol. 15, no 1, p. 23-42Article in journal (Refereed) Published
Abstract [en]

To identify series of residential burglaries, detecting linked crimes performed by the same constellations of criminals is necessary. Comparison of crime reports today is difficult as crime reports traditionally have been written as unstructured text and often lack a common information-basis. Based on a novel process for collecting structured crime scene information, the present study investigates the use of clustering algorithms to group similar crime reports based on combined crime characteristics from the structured form. Clustering quality is measured using Connectivity and Silhouette index (SI), stability using Jaccard index, and accuracy is measured using Rand index (RI) and a Series Rand index (SRI). The performance of clustering using combined characteristics was compared with spatial characteristic. The results suggest that the combined characteristics perform better or similar to the spatial characteristic. In terms of practical significance, the presented clustering approach is capable of clustering cases using a broader decision basis.

Place, publisher, year, edition, pages
World Scientific, 2016
Keywords
Crime clustering, residential burglary analysis, decision support system, combined distance metric
National Category
Computer Sciences
Identifiers
urn:nbn:se:bth-11779 (URN)10.1142/S0219622015500339 (DOI)000371127600003 ()
Available from: 2016-04-01 Created: 2016-04-01 Last updated: 2018-01-10Bibliographically approved
Organisations
Identifiers
ORCID iD: ORCID iD iconorcid.org/0000-0002-9316-4842

Search in DiVA

Show all publications