Neesha Kodagoda

INVISQUE: intuitive information exploration through interactive visualization

In this paper we present INVISQUE, a novel system designed for interactive information exploration. Instead of a conventional list-style arrangement, in INVISQUE information is represented by a two-dimensional spatial canvas, with each dimension representing user-defined semantics. Search results are presented as index cards, ordered in both dimensions. Intuitive interactions are used to perform tasks such as keyword searching, results browsing, categorizing, and linking to online resources such as Google and Twitter. The interaction-based query style also naturally lends the system to different types of user input such as multi-touch gestures. As a result, INVISQUE gives users a much more intuitive and smooth experience of exploring large information spaces.

Human-centered computing in international development

This workshop continues the dialog on exploring the challenges in applying, extending, and inventing appropriate methods and contributions of Humancentered Computing (HCC) to International economic and community development, borne out of tremendously successful HCI4D workshops at CHI 2007 and 2008. The workshop aims at 1) providing a platform to discuss interaction design practices that allow for meaningful embedding of interactive systems in the cultural, infrastructural, and political settings where they will be used 2) addressing interaction design issues in developing regions, as well as areas in the developed world marginalized by poverty or other barriers. We hope to continue to extend the boundaries of the field of Humancentered Computing (HCC) by spurring on more discussion on how existing methods and practices can be adapted/ modified, and how new practices be developed, to combat

Interactive visualization for low literacy users: from lessons learnt to design

This paper aims to address the problems low literacy (LL) users face when searching for information online. The first part of this paper summarizes the problems that LL users face, and establishes a set of design principles for interfaces suitable for LL users. This is followed by a description of how these design principles are mapped to a novel interface for interactive data retrieval. The interface was realized into a working system and evaluated against a traditional web interface for both high literacy (HL) and LL users. The suitability of the designs was analyzed using performance data, subjective feedback and an observational analysis. The findings from the study suggest that LL users perform better and prefer the proposed designs over a traditional web interface.

How Analysts Think Anchoring, Laddering and Associations

In this paper we present our observations of how seven criminal intelligence analysts use inference making and storytelling to create explanations. Adapting the Critical Decision Method, we observe they engage in a process of anchoring to gain traction and initiate further inquires; they engage in a laddering process, where they develop explanations to extend or elaborate their ideas that create bridges to new understanding; and they complement their anchoring and laddering activities by associative questioning to discover what other associations exists. These un-intended associations can lead to insights, or generate new patterns recognisable by intuition. We envisage a better understanding of how analyst think and reason might help design software that encourages insight.

How Analysts Think Intuition, Leap of Faith and Insight

In this paper we present three cognitive acts analysts use while solving criminal cases: intuition, leap of faith and insight. We used the Emergent Themes Analysis to find out how these cognitive acts help analysts’ inference making and what are their features. We analyzed the interviews with six analysts that were done with the Critical Decision Method. What we discovered is that intuition, leap of faith and insight relate to each other creating an integrated process while solving problems. We propose that a leap of faith occurs between intuition and insight. This is a preliminary study that we plan to develop further. However, current results are interesting enough that we hope the findings may help us design computer systems that will facilitate the process of solving criminal cases by analysts.Criminal intelligence analysis, analytical reasoning, inference making, intuition, leap of faith, insight, critical decision method.

How analysts think (?): inference making strategies

In this paper we present early observations of how seven criminal intelligence analysts think and how the make inferences. We used the Critical Decision Method to identify the causal mechanisms of how they think and reason, i.e. how they organize, structure and assemble their information, understandings and inferences. We envisaged that this would enable us to design software to support the structuring of arguments and the evidential reasoning process. Our early observations suggest that analytic reasoning is not straight-forward, but appears chaotic and haphazard, and sometimes cyclic; and that inference making – abduction, induction and deduction – are not independent processes, but are closely intertwined. These processes interact dynamically, each producing outcomes that become anchors used by the others.

Using Machine Learning to Infer Reasoning Provenance From User Interaction Log Data: Based on the Data/Frame Theory of Sensemaking

The reconstruction of analysts’ reasoning processes (reasoning provenance) during complex sensemaking tasks can support reflection and decision making. One potential approach to such reconstruction is to automatically infer reasoning from low-level user interaction logs. We explore a novel method for doing this using machine learning. Two user studies were conducted in which participants performed similar intelligence analysis tasks. In one study, participants used a standard web browser and word processor; in the other, they used a system called INVISQUE (Interactive Visual Search and Query Environment). Interaction logs were manually coded for cognitive actions based on captured think-aloud protocol and posttask interviews based on Klein, Phillips, Rall, and Pelusos’s data/frame model of sensemaking as a conceptual framework. This analysis was then used to train an interaction frame mapper, which employed multiple machine learning models to learn relationships between the interaction logs and the codings. Our results show that, for one study at least, classification accuracy was significantly better than chance and compared reasonably to a reported manual provenance reconstruction method. We discuss our results in terms of variations in feature sets from the two studies and what this means for the development of the method for provenance capture and the evaluation of sensemaking systems.

Trialling the SMART Approach Identifying and Assessing Sense-making

It is important to develop tools that support sense-making by providing representations that help to capture the externalisation of the thinking process. The paper proposes SMART, a new method for identifying the sense-making processes of experts by combing probes with cognitive task analysis methods. The Data-Frame sense-making model is used as a theoretical frame, and the probes have been developed around the model to elicit experts’ sense-making strategies. However, we found that the SMART probes presently lacked the resolution to capture the experts sense-making and a stronger emphasis of cognitive task analysis methods and observations were required to interpret the findings.

Information seeking behaviour model as a theoretical lens: high and low literate users behaviour process analysed

Motivation -- The paper focuses on how information seeking behaviour model is used as a theoretical lens to analyse high and low literate users online behaviour which in turn will support interface design suggestions. Research approach -- Five high and five low literate users of a local charity which provides social service information participated to carry out four online information seeking tasks. Data were captured using think-aloud, video, observation and semi structured interview techniques. A data analysis on the study previously discovered eight information seeking behaviour strategies: Reading, Scanning, Focus, Satisfied, Verification, Recovery, Trajectories, Representation and Abandon. Several information seeking behaviour models were evaluated prior to selecting Ellis (1989) information seeking behaviour model which includes features such as: starting, chaining, browsing, differentiating, monitoring, extracting, verifying, and ending. The model is used as a theoretical lens to analyse the data combining with the previous findings to make interface design suggestions. The study will not validate the correctness or the features of Ellis model. Findings/Design -- The analysis uncovered two variations of Ellis model for the high and low literate users, and how the models were used to give interface design suggestions. Research limitations/Implications -- The small sample size of five high and five low literate participants, limited the possibility of generalizing the findings. Originality/Value -- The low and high literate users information seeking behaviour were analysed using Ellis model as a theoretical lens along with the previously identified information seeking behaviour strategies of these users. These finds of the refined models are used to suggest interface design to improve the low literate users online information seeking. Take away message -- The models will be used to suggest interface design recommend for low literate users. We hope the design suggestions will help improve the low literate users online information seeking.

How analysts think: sense-making strategies in the analysis of temporal evolution and criminal network structures and activities

Analysis of criminal activity based on offenders’ social networks is an established procedure in intelligence analysis. The complexity of the data poses an obstacle for analysts to gauge network developments, e.g. detect emerging problems. Visualization is a powerful tool to achieve this, but it is essential to know how the analysts’ sense-making strategies can be supported most efficiently. Based on a think aloud study we identified ten cognitive strategies on a general level to be useful for designers. We also provide some examples how these strategies can be supported through appropriate visualizations.