liu.seSearch for publications in DiVA
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • oxford
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Data Abstraction and Pattern Identification in Time-series Data
Linköping University, Department of Science and Technology, Media and Information Technology. Linköping University, Faculty of Science & Engineering.
2019 (English)Doctoral thesis, comprehensive summary (Other academic)
Abstract [en]

Data sources such as simulations, sensor networks across many application domains generate large volumes of time-series data which exhibit characteristics that evolve over time. Visual data analysis methods can help us in exploring and understanding the underlying patterns present in time-series data but, due to their ever-increasing size, the visual data analysis process can become complex. Large data sets can be handled using data abstraction techniques by transforming the raw data into a simpler format while, at the same time, preserving significant features that are important for the user. When dealing with time-series data, abstraction techniques should also take into account the underlying temporal characteristics.  

This thesis focuses on different data abstraction and pattern identification methods particularly in the cases of large 1D time-series and 2D spatio-temporal time-series data which exhibit spatiotemporal discontinuity. Based on the dimensionality and characteristics of the data, this thesis proposes a variety of efficient data-adaptive and user-controlled data abstraction methods that transform the raw data into a symbol sequence. The transformation of raw time-series into a symbol sequence can act as input to different sequence analysis methods from data mining and machine learning communities to identify interesting patterns of user behavior.  

In the case of very long duration 1D time-series, locally adaptive and user-controlled data approximation methods were presented to simplify the data, while at the same time retaining the perceptually important features. The simplified data were converted into a symbol sequence and a sketch-based pattern identification was then used to identify patterns in the symbolic data using regular expression based pattern matching. The method was applied to financial time-series and patterns such as head-and-shoulders, double and triple-top patterns were identified using hand drawn sketches in an interactive manner. Through data smoothing, the data approximation step also enables visualization of inherent patterns in the time-series representation while at the same time retaining perceptually important points.  

Very long duration 2D spatio-temporal eye tracking data sets that exhibit spatio-temporal discontinuity was transformed into symbolic data using scalable clustering and hierarchical cluster merging processes, each of which can be parallelized. The raw data is transformed into a symbol sequence with each symbol representing a region of interest in the eye gaze data. The identified regions of interest can also be displayed in a Space-Time Cube (STC) that captures both the temporal and contextual information. Through interactive filtering, zooming and geometric transformation, the STC representation along with linked views enables interactive data exploration. Using different sequence analysis methods, the symbol sequences are analyzed further to identify temporal patterns in the data set. Data collected from air traffic control officers from the domain of Air traffic control were used as application examples to demonstrate the results.

Place, publisher, year, edition, pages
Linköping: Linköping University Electronic Press, 2019. , p. 58
Series
Linköping Studies in Science and Technology. Dissertations, ISSN 0345-7524 ; 2030
National Category
Media Engineering
Identifiers
URN: urn:nbn:se:liu:diva-162220DOI: 10.3384/diss.diva-162220ISBN: 9789179299651 (print)OAI: oai:DiVA.org:liu-162220DiVA, id: diva2:1372703
Public defence
2019-12-13, Domteatern, Visualiseringscenter C, Kungsgatan 54, 60233 Norrköping, Norrköping, 09:15 (English)
Opponent
Supervisors
Available from: 2019-11-25 Created: 2019-11-25 Last updated: 2019-11-25Bibliographically approved
List of papers
1. Shape Grammar Extraction for Efficient Query-by-Sketch Pattern Matching in Long Time Series
Open this publication in new window or tab >>Shape Grammar Extraction for Efficient Query-by-Sketch Pattern Matching in Long Time Series
2016 (English)Conference paper, Published paper (Refereed)
Abstract [en]

Long time-series, involving thousands or even millions of time steps, are common in many application domains but remain very difficult to explore interactively. Often the analytical task in such data is to identify specific patterns, but this is a very complex and computationally difficult problem and so focusing the search in order to only identify interesting patterns is a common solution. We propose an efficient method for exploring user-sketched patterns, incorporating the domain expert’s knowledge, in time series data through a shape grammar based approach. The shape grammar is extracted from the time series by considering the data as a combination of basic elementary shapes positioned across different am- plitudes. We represent these basic shapes using a ratio value, perform binning on ratio values and apply a symbolic approximation. Our proposed method for pattern matching is amplitude-, scale- and translation-invariant and, since the pattern search and pattern con- straint relaxation happen at the symbolic level, is very efficient permitting its use in a real-time/online system. We demonstrate the effectiveness of our method in a case study on stock market data although it is applicable to any numeric time series data.

Place, publisher, year, edition, pages
Institute of Electrical and Electronics Engineers (IEEE), 2016. p. 10
Series
IEEE Conference on Visual Analytics Science and Technology, ISSN 2325-9442
Keywords
User-queries, Sketching, Time Series, Symbolic ap-proximation, Regular Expression, Shape Grammar
National Category
Engineering and Technology Computer Sciences Computer Systems Computer Vision and Robotics (Autonomous Systems) Bioinformatics (Computational Biology)
Identifiers
urn:nbn:se:liu:diva-134334 (URN)10.1109/VAST.2016.7883518 (DOI)000402056500013 ()978-1-5090-5661-3 (ISBN)
Conference
2016 IEEE CONFERENCE ON VISUAL ANALYTICS SCIENCE AND TECHNOLOGY (VAST), October 23-28, Baltimore, USA
Funder
Swedish Research Council, 2013-4939
Available from: 2017-02-03 Created: 2017-02-03 Last updated: 2019-11-25Bibliographically approved
2. Supporting Exploration of Eye Tracking Data: Identifying Changing Behaviour Over Long Durations
Open this publication in new window or tab >>Supporting Exploration of Eye Tracking Data: Identifying Changing Behaviour Over Long Durations
Show others...
2016 (English)In: BEYOND TIME AND ERRORS: NOVEL EVALUATION METHODS FOR VISUALIZATION, BELIV 2016, ASSOC COMPUTING MACHINERY , 2016, p. 70-77Conference paper, Published paper (Refereed)
Abstract [en]

Visual analytics of eye tracking data is a common tool for evaluation studies across diverse fields. In this position paper we propose a novel user-driven interactive data exploration tool for understanding the characteristics of eye gaze movements and the changes in these behaviours over time. Eye tracking experiments generate multidimensional scan path data with sequential information. Many mathematical methods in the past have analysed one or a few of the attributes of the scan path data and derived attributes such as Area of Interest (AoI), statistical measures, geometry, domain specific features etc. In our work we are interested in visual analytics of one of the derived attributes of sequential data-the: AoI and the sequences of visits to these AoIs over time. In the case of static stimuli, such as images, or dynamic stimuli, like videos, having predefined or fixed AoIs is not an efficient way of analysing scan path patterns. The AoI of a user over a stimulus may evolve over time and hence determining the AoIs dynamically through temporal clustering could be a better method for analysing the eye gaze patterns. In this work we primarily focus on the challenges in analysis and visualization of the temporal evolution of AoIs. This paper discusses the existing methods, their shortcomings and scope for improvement by adopting visual analytics methods for event-based temporal data to the analysis of eye tracking data.

Place, publisher, year, edition, pages
ASSOC COMPUTING MACHINERY, 2016
Keywords
Eye tracking; pattern analysis; scan path; time evolving AoIs; Clustering of Fixations; ActiviTree
National Category
Media Engineering
Identifiers
urn:nbn:se:liu:diva-133129 (URN)10.1145/2993901.2993905 (DOI)000387865100009 ()978-1-4503-4818-8 (ISBN)
Conference
6th Bi-Annual Workshop (BELIV)
Available from: 2016-12-12 Created: 2016-12-09 Last updated: 2019-11-25
3. Identification of Temporally Varying Areas of Interest in Long-Duration Eye-Tracking Data Sets
Open this publication in new window or tab >>Identification of Temporally Varying Areas of Interest in Long-Duration Eye-Tracking Data Sets
Show others...
2019 (English)In: IEEE Transactions on Visualization and Computer Graphics, ISSN 1077-2626, E-ISSN 1941-0506, p. 87-97Article in journal (Refereed) Published
Abstract [en]

Eye-tracking has become an invaluable tool for the analysis of working practices in many technological fields of activity. Typically studies focus on short tasks and use static expected areas of interest (AoI) in the display to explore subjects’ behaviour, making the analyst’s task quite straightforward. In long-duration studies, where the observations may last several hours over a complete work session, the AoIs may change over time in response to altering workload, emergencies or other variables making the analysis more difficult. This work puts forward a novel method to automatically identify spatial AoIs changing over time through a combination of clustering and cluster merging in the temporal domain. A visual analysis system based on the proposed methods is also presented. Finally, we illustrate our approach within the domain of air traffic control, a complex task sensitive to prevailing conditions over long durations, though it is applicable to other domains such as monitoring of complex systems. 

Place, publisher, year, edition, pages
Institute of Electrical and Electronics Engineers (IEEE), 2019
Keywords
Eye-tracking data, areas of interest, clustering, minimum spanning tree, temporal data, spatio-temporal data
National Category
Computer Systems
Identifiers
urn:nbn:se:liu:diva-152714 (URN)10.1109/TVCG.2018.2865042 (DOI)000452640000009 ()30183636 (PubMedID)2-s2.0-85052788669 (Scopus ID)
Note

Funding agencies: Swedish Research Council [2013-4939]; RESKILL project - Swedish Transport Administration; Swedish Maritime Administration; Swedish Air Navigation Service Provider LFV

Available from: 2018-11-16 Created: 2018-11-16 Last updated: 2019-11-25Bibliographically approved
4. Analysis of Long Duration Eye-Tracking Experiments in a Remote Tower Environment
Open this publication in new window or tab >>Analysis of Long Duration Eye-Tracking Experiments in a Remote Tower Environment
Show others...
2019 (English)Conference paper, Published paper (Refereed)
Abstract [en]

Eye-Tracking experiments have proven to be of great assistance in understanding human computer interaction across many fields. Most eye-tracking experiments are non-intrusive and so do not affect the behaviour of the subject. Such experiments usually last for just a few minutes and so the spatio- temporal data generated by the eye-tracker is quite easy to analyze using simple visualization techniques such as heat maps and animation. Eye tracking experiments in air traffic control, or maritime or driving simulators can, however, last for several hours and the analysis of such long duration data becomes much more complex. We have developed an analysis pipeline, where we identify visual spatial areas of attention over a user interface using clustering and hierarchical cluster merging techniques. We have tested this technique on eye tracking datasets generated by air traffic controllers working with Swedish air navigation services, where each eye tracking experiment lasted for ∼90 minutes. We found that our method is interactive and effective in identification of interesting patterns of visual attention that would have been very difficult to locate using manual analysis.

Keywords
Remote tower, Eye tracking, Spatio-temporal clustering
National Category
Media Engineering
Identifiers
urn:nbn:se:liu:diva-160959 (URN)
Conference
Thirteenth USA/Europe Air Traffic Management Research and Development Seminar (ATM2019), Vienna, Austria, June 17-21, 2019
Funder
Swedish Transport AdministrationSwedish Research Council
Available from: 2019-10-16 Created: 2019-10-16 Last updated: 2019-11-25Bibliographically approved
5. Comparison of Attention Behaviour Across User Sets through Automatic Identification of Common Areas of Interest
Open this publication in new window or tab >>Comparison of Attention Behaviour Across User Sets through Automatic Identification of Common Areas of Interest
Show others...
2019 (English)Conference paper, Published paper (Refereed)
Abstract [en]

Eye tracking is used to analyze and compare user behaviour within numerous domains, but long duration eye tracking experiments across multiple users generate millions of eye gaze samples, making the data analysis process complex. Usually the samples are labelled into Areas of Interest (AoI) or Objects of Interest (OoI), where the AoI approach aims to understand how a user monitors different regions of a scene while OoI identification uncovers distinct objects in the scene that attract user attention. Using scalable clustering and cluster merging techniques that require minimal user input, we label AoIs across multiple users in long duration eye tracking experiments. Using the common AoI labels then allows direct comparison of the users as well as the use of such methods as Hidden Markov Models and Sequence mining to uncover common and distinct behaviour between the users which, until now, has been prohibitively difficult to achieve.

National Category
Engineering and Technology
Identifiers
urn:nbn:se:liu:diva-161999 (URN)
Conference
Hawaii International Conference on System Sciences
Available from: 2019-11-15 Created: 2019-11-15 Last updated: 2019-11-25

Open Access in DiVA

fulltext(14719 kB)31 downloads
File information
File name FULLTEXT02.pdfFile size 14719 kBChecksum SHA-512
a98feaab9be11827461ca27938c44b3429f8a4d6ecef8e77d38d6599d0123b221ec01ad78d531cc0916352bc9fc2034d9210f53d411dbacefd587a8486692aa0
Type fulltextMimetype application/pdf
Order online >>

Other links

Publisher's full text

Authority records BETA

Muthumanickam, Prithiviraj

Search in DiVA

By author/editor
Muthumanickam, Prithiviraj
By organisation
Media and Information TechnologyFaculty of Science & Engineering
Media Engineering

Search outside of DiVA

GoogleGoogle Scholar
Total: 31 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

doi
isbn
urn-nbn

Altmetric score

doi
isbn
urn-nbn
Total: 88 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • oxford
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf