Katerina Vrotsou

ActiviTree: Interactive Visual Exploration of Sequences in Event-Based Data Using Graph Similarity

The identification of significant sequences in large and complex event-based temporal data is a challenging problem with applications in many areas of todays information intensive society. Pure visual representations can be used for the analysis, but are constrained to small data sets. Algorithmic search mechanisms used for larger data sets become expensive as the data size increases and typically focus on frequency of occurrence to reduce the computational complexity, often overlooking important infrequent sequences and outliers. In this paper we introduce an interactive visual data mining approach based on an adaptation of techniques developed for Web searching, combined with an intuitive visual interface, to facilitate user-centred exploration of the data and identification of sequences significant to that user. The search algorithm used in the exploration executes in negligible time, even for large data, and so no pre-processing of the selected data is required, making this a completely interactive experience for the user. Our particular application area is social science diary data but the technique is applicable across many other disciplines.

Everyday Life Discoveries: Mining and Visualizing Activity Patterns in Social Science Diary Data

The ability to identify and examine patterns of activities is a key tool for social and behavioural science. In the past this has been done by statistical or purely visual methods but automated sequential pattern analysis through sophisticated data mining and visualization tools for pattern location and evaluation can open up new possibilities for interactive exploration of the data. This paper describes the addition of a sequential pattern identification method to the visual activity-analysis tool, VISUAL-TimePAcTS, and its effectiveness in the process of pattern analysis in social science diary data. The results have shown that the method correctly identifies patterns and conveys them effectively to the social scientist in a manner that allows them quick and easy understanding of the significance of the patterns.

2D and 3D representations for feature recognition in time geographical diary data

Time geographical representations are becoming a common approach to analysing spatio-temporal data. Such representations appear intuitive in the process of identifying patterns and features as paths of populations form tracks through the 3D space, which can be seen converging and diverging over time. In this article, we compare 2D and 3D representations within a time geographical visual analysis tool for activity diary data. We identify a representative task and evaluate task performance between the two representations. The results show that the 3D representation has benefits over the 2D representation for feature identification but also indicate that these benefits can be lost if the 3D representation is not carefully constructed to help the user to see them.

SimpliFly: A Methodology for Simplification and Thematic Enhancement of Trajectories

Movement data sets collected using todays advanced tracking devices consist of complex trajectories in terms of length, shape, and number of recorded positions. Multiple additional attributes characterizing the movement and its environment are often also included making the level of complexity even higher. Simplification of trajectories can improve the visibility of relevant information by reducing less relevant details while maintaining important movement patterns. We propose a systematic stepwise methodology for simplifying and thematically enhancing trajectories in order to support their visual analysis. The methodology is applied iteratively and is composed of: (a) a simplification step applied to reduce the morphological complexity of the trajectories, (b) a thematic enhancement step which aims at accentuating patterns of movement, and (c) the representation and interactive exploration of the results in order to make interpretations of the findings and further refinement to the simplification and enhancement process. We illustrate our methodology through an analysis example of two different types of tracks, aircraft and pedestrian movement.

Exploring city structure from georeferenced photos using graph centrality measures

We explore the potential of applying graph theory measures of centrality to the network of movements extracted from sequences of georeferenced photo captures in order to identify interesting places and explore city structure. We adopt a systematic procedure composed of a series of stages involving the combination of computational methods and interactive visual analytics techniques. The approach is demonstrated using a collection of Flickr photos from the Seattle metropolitan area.

Are we what we do? Exploring group behaviour through user-defined event-sequence similarity

The study of human activity in space and time is an inherent part of human geography. In order to perform such studies, data on the time use of individuals, in terms of sequence and timing of performed activities, are collected and analysed. A common assumption when analysing individuals’ time use is that groups that exhibit similar background and demographic characteristics also display similarities in how they use their time to structure their daily lives. In this article, we set out to investigate the correctness of such assumptions. We propose a visual analytics process based on sequence similarity measures tailored to event-based data such as performed activity sequences. The process allows an analyst to retrieve similarly behaving records according to user-selected similarity preferences and interactively explore aspects of this similarity in a multiple linked-view environment.

A qualitative study of similarity measures in event-based data

This paper presents an interview-based study of the definition of sequence similarity in different application areas of event-based data. The applicability of nine identified measures across these areas is investigated and discussed. The work helps highlight what are the core characteristics sought when analysing event-based data and performs a first validation of this across disciplines. The results of the study make a solid basis for follow-up evaluations of the practical applicability and usability of the similarity measures.

Women's mobility and the situational conditions of rape: cases reported to hospitals

A third of all rapes in Stockholm, the capital of Sweden, take place in public outdoor places. Yet, little is known about the events that precede this type of sexual offence and less about the situational context of rape. This study aims to improve the understanding of the nature of situational conditions that immediately precede events of rape. Using medical records of 147 rape victims during 2012 and 2013, we constructed time- and place-specific records of the places women traveled through or spent time at, the activities they engaged in, and the people they interacted with sequentially over the course of the day when they were raped. The analysis uses visualization tools (VISUAL-TimePAcTS), Geographical Information Systems, and conditional logistic regression to identify place-, context-, and social interaction–related factors associated with the onset of rape. Results for this sample of cases reported to hospitals show that being outdoors was not necessarily riskier for women when compared with indoor public settings; some outdoor environments were actually protective, such as streets. Being in a risky social context and engaging in a risky activity before the event was associated with an increased risk of rape, and the risk escalated over the day. Among those women who never drank alcohol, the results were similar to what was observed in the overall sample, which suggests that risky social interaction and risky activity made independent contributions to the risk of rape. The article finishes with suggestions for rape prevention.

KD-photomap: Exploring photographs in space and time

KD-photomap is a web-based visual analytics system for browsing collections of geotagged Flickr photographs in search of interesting pictures, places, and events. Spatial filtering of the data is performed through zooming, moving or searching along the map. Temporal filtering is possible through defining time windows using interactive histograms and calendar controls. Information about the number and spatiotemporal distribution of photos captured in an explored area is continuously provided using various visual cues.

Scalable Cluster Analysis of Spatial Events

Clustering of massive data is an important analysis tool but also challenging since the data often does not fit in RAM. Many clustering algorithms are thus severely memory-bound. This paper proposes a deterministic density clustering algorithm based on DBSCAN that allows to discover arbitrary shaped clusters of spatio-temporal events that (1) achieves scalability to very large datasets not fitting in RAM and (2) exhibits significant execution time improvements for processing the full dataset compared to plain DBSCAN. The proposed algorithms integration with interactive visualization methods allows for visual inspection of clustering results in the context of the analysis task; several alternatives are discussed by means of an application example about traffic data analysis.