WikiISI/Visualization/HCI

From AI Wiki

Visualization/HCI in ISI

Abstract

Human-computer interaction (HCI) is the study of interaction between people (users) and computers. This encompasses not only ease of use, but also new interaction techniques for supporting user tasks, providing better access to information, and creating more powerful forms of communication. Security and law enforcement organizations have access to extremely large volumes of data, and powerful intelligence tools. However, what is needed is the analysis of this data, and the judgment of experts based on this analysis. Visualization can play a very important role in this analysis and decision-making process. In the following sections we will discuss in greater detail the need for visualization in ISI, current applications, and current visualization tools for ISI.

Introduction

Although visualization is a relatively new research area, visualization has a long history. The first known maps appeared in the 12th century^[1], while multidimensional representations appeared in the 19th century(Tufte, 1983)

In the field of science, Bertin identified the basic elements of diagrams in 1967^[2]. Most early visualization research focused on statistical graphs ^[3]. With the data explosion in 1980s^[4], NSF launched the “Scientific visualization” initiative in 1985, and the first IEEE visualization conference was held in 1990.

In nonscientific contexts, the term “information visualization” was first used in Robertson et al. (1989). Early information visualization systems emphasized interactivity and animation^[5], interfaces to support dynamic queries^[6], and layout algorithms^[7]. Later visualization systems emphasized subject hierarchy of the Internet^[8], summarizing the contents of a document^[9], describing online behaviors^[10], displaying website usage patterns^[11], and visualizing the structures of a knowledge domain^[12]. In this chapter, we will focus on current research and trends in the field of visualization for information and security informatics.

Literature Review

A Theoretical Foundation for Visualization

The human eye can process many visual cues simultaneously^[13]. People have a remarkable ability to recall pictorial images (Standing et al., 1970). Visuals aid people to find patterns. But patterns will be invisible if they are not presented in certain ways. Understanding visual perception can be helpful in the design of visualization system.

Different parts of human memory can be enhanced by visualization in different ways^[13]:

• Iconic memory is the memory buffer where pre-attentive processing operates. According to the visual processing channel theory^[13], certain visual patterns can be detected at this stage without having to go through the cognition process. Hence we can design effective visualizations that rely on understanding the perception of patterns.
• Working memory integrates information from iconic memory and long-term memory for problem solving. Patterns perceived by pre-attentive processing are mapped into patterns of the information space. Visualization can serve as an external memory, saving space in the working memory.
• Long-term memory stores information in a network of linked concepts ^[14][15]. Using proximity to represent relationships among concepts in constructing a concept map has a long history. Visualization also uses proximity to indicate semantic relationships among concepts.

A Framework for Information Visualization

Effective and efficient knowledge discovery on the web has the following components:

• Identification of relevant information
• Visualization of relevant information
• Manipulation of the information

Shneiderman (1996)^[6] classified seven tasks that cover information browsing and searching: Overview, Zoom, Filter, Details-on-demand, Relate, History, and Extract. Visual interfaces that support these seven tasks are a potential solution to information overload, vocabulary difference, and cognitive content mapping^{[2][16][17][18]}. Shneiderman (1996) also classified types of representation methods: 1-D, 2-D, 3-D, Multi-dimensional, Temporal, Tree, and Network.

• 1-D Visualization: This technique represents information as one-dimensional visual objects in a linear^[11][9] or a circular^[19] manner. It can be used to display contents of a single document (Hearst, 1995; Salton et al., 1995), or to provide an overview of a document collection^[11]. Colors usually represent some attributes, and a second axis may also play a role.
• 2-D/Map Visualization: This technique represents information as two-dimensional visual objects to help user deal with the large number of categories created for mass textual data. This can be seen in COPLINK's STV, and in i2 Analyst's Notebook.

Figure 1. System Design for relational parser

• 3-D Visualization: This technique represents information as three-dimensional visual objects. This type of representation can be seen in Starlight.

Figure 2.

• Multidimensional Visualization: This technique represents information as multidimensional objects and projects them into a three-dimensional or a two-dimensional space. We can see this type of visualization in the multidimensional view in Starlight.

Figure 3.

• Tree Visualization: This technique can be used to represent hierarchical relationships. A major challenge is the exponential growth of nodes, and different layout algorithms have been applied to deal with this problem. One example is the tree layout in i2 Analyst's Notebook.

Figure 4.

• Network Visualization: This technique is used to represent complex relationships that cannot be represented by a simple tree structure. This can be seen in COPLINK's CAN (Criminal Activity Network), which has been used to study gang networks, etc.

Figure 5. COPLINK: A 57-member Gang network: Nodes represent individual criminals labeled by their names. Links represent relationships between criminals.

• Temporal Visualization: This technique can be used to information based on temporal order. Location and animation are commonly used visual variables to reveal the temporal aspect of information. An example is COPLINK STV, which is discussed in greater detail in the case study.

Figure 6. COPLINK STV

A visualization system usually applies several methods at the same time.

Evaluation Research for Information Visualization

There are two main approaches commonly used to study the validity of information visualization:

• Empirical usability studies attempt to understand the pros and cons of specific visualization designs or systems. Either a laboratory experiments approach or a de-featuring approach is commonly used. Complex, realistic, task-driven evaluation studies have been conducted frequently. They could measure usefulness. But it is difficult to identify each visualization factors’ contribution. Behavioral methods also need to be considered.
• Fundamental perception studies and theory building attempt to investigate basic perceptual effects of certain visualization factors or stimuli. Theories from psychology and neuroscience are used to understand the perceptual impact of visualization parameters as animation, information density, 3-D effect and combinations of visual cues. However, results may be applied only to the particular visualization system under study.

Visualization for Various ISI Task Categories

Intelligence Analysis and Knowledge Discovery

Within this domain, there are several applications for visualization. Abbasi and Chen (2006)^[20] developed an authorship visualization called Writeprints that can help identify individuals based on their writing style. This can help create greater user accountability and thus prevent misuse of online anonymity. Deligiannidis et al. (2006) ^[21] have created a 3-D visualization tool called Semantic Analytics Visualization, which can address a variety of issues such as aviation safety, provenance and trust of data sources, and the document-access problem of insider threats. Buennemeyer et al. (2006) ^[22] developed a prototypical Gigapixel Intelligence Analysis Navigation Tool (GIANT) to aid in Intelligence Analysis using high resolution displays. COPLINK, developed at the AI Lab at the University of Arizona, is another widely-used intelligence analysis and knowledge discovery tool for law enforcement. We will discuss COPLINK in greater detail in the Case Studies section.

Access Control, Privacy, and Cyber Trust

Abbasi and Chen's Writeprints^[20] has also been used as a visualization tool that addresses the issue of online trust.

Surveillance and Emergency Response

In response to the critical need of early detection of potential infectious disease outbreaks or bioterrorism events, public health syndromic surveillance systems have been rapidly developed and deployed in recent years^[23]. These systems rely on visualizations to facilitate interactive data exploration in order to aid in the decision making process. BioPortal, a syndromic surveillance application developed at the AI Lab at the University of Arizona, has an advanced visualization module called the Spatial Temporal Visualizer (STV). RODS, developed at the University of Pittsburgh, employs a GIS module to depict data spatially.

Infrastructure Protection and Cyber Security

Computer network security has become an increasingly pressing issue for many organizations. Starlight, built by Battelle at the Pacific Northwest National Laboratory (PNNL), is a tool that can enable network analysts to quickly achieve and maintain an in-depth understanding of network vulnerabilities and security status. We review Starlight in greater detail in the Case Study section.

Terrorism Informatics and Countermeasures

Analysis of terrorist social networks is essential for developing effective strategies against terrorism. Such networks can be visualized using social network analysis tools. Several researchers have explored this area. Yang et al. (2006) ^[24] explored fisheye views and fractal views to interactively study large and complex networks. Reid and Chen (2006) ^[25] applied domain visualization techniques such as content map analysis, block-modeling, and co-citation analysis to identify the core researchers and knowledge creation approaches in terrorism.

Emerging Applications

There are several emergent application areas that use visualization tools for intelligence and security informatics. Here we will briefly summarize some of them.

Intelligent Face Recognition: Such systems use biometrics technology, and can be used for national and international security^[26].

Systems for Situational Awareness: Enhanced automated situational awareness can be a crucial component in developing intelligent infrastructures for safer environments. Vision-based analysis of people and vehicles can be one direction for emerging applications to explore ^[27].

Image Retrieval and Image Matching: These applications can be used in homeland security tasks such as human tracking or matching photos of dead against missing individuals ^[28].

Research/System Design

Most visualization systems for ISI have a basic common architecture. After data is collected from several sources, it is cleaned to remove noise, parsed and integrated. Relevant features are then extracted from the dataset. Then, information analysis is applied in order to generate data for visualization, and generated data is then loaded into the visualization tool in order to facilitate analysis by an expert.

Figure 7. ISI Visualization Systems

Data Preparation

During the Data Preparation stage, data in different forms are collected from various sources. After that, useful data need to be extracted, parsed and integrated from raw data collected.

Indexing

In the second stage, the extracted data is indexed, using one of three techniques:

• Semantic information may be extracted.
• Automatic indexing may be used to represent the content of each document as a vector of key terms. During this process, natural language processing noun-phrasing technique can capture a rich linguistic representation of document content. Most noun phrasing techniques rely on a combination of part-of-speech-tagging (POST) and grammatical phrase-forming rules.
• Machine learning may be combined with either a rule-based approach or a statistical approach in order to extract entities from textual documents.

Information Analysis

In this stage, the system prepares the data for visualization. Dimensionality reduction is important for data preparation when dealing with high dimensional data. Data should be 1-D, 2-D or 3-D for visualization, so dimensionality reduction is applied to data to fit data into low dimensional space. Classification or clustering, two common information analysis methods, are also applied to generate data for visualization.

Visualization

Data generated in the last stage is then loaded into the visualization tool to show the final visual effects. Different visualization tools have different cues, functions and visual effects which can be used to facilitate various analysis needs of experts.

Case Studies

In this section, we present three systems, COPLINK, i2 Analyst's Notebook and Starlight, and their visualization functionalities. These three systems are widely used, in the US and abroad, by several intelligence and law enforcement agencies.

Cast Study 1: COPLINK

Introduction

The COPLINK system was initially developed by the University of Arizona Artificial Intelligence Lab with funding from the National Institute of Justice and the National Science Foundation since 1997. With additional venture funding and product development, Knowledge Computing Corporation (KCC) currently distributes, maintains, and updates the commercially available COPLINK Solution Suite. It provides cross-jurisdictional information sharing, analysis, visualization and research for the law enforcement and intelligence community for border and national security.

COPLINK Solution Suite has already been used by police officers around the country and has a lot of successful stories in California, Arizona, Florida and etc^[29].

Overview

COPLINK employs several different 2-D visualization techniques, the data representation methods used in COPLINK include tree, spatial, temporal and network.

• COPLINK Hyperbolic Tree Visualization

This visualization tool utilizes a tree structure, and supports all seven tasks as defined by Shneiderman, namely Overview, Zoom, Filter, Details-on-demand, Relate, History, and Extract. In the hyperbolic tree view of associations in COPLINK, an officer can search for all entities related to a suspect.

The thickness of the arcs indicates the weight (closeness) of the relationship. The color indicates the entity type. These types can be seen at the bottom of Fig. 4 (green: person, pink: address, brown: vehicle, black: crime type, blue: organization). Terms may also be selected from the hyperbolic tree using the mouse button. The screen shown in Fig. 3 shows an example where the figure is displayed showing all entities associated with each of these search terms.

Figure 8. COPLINK hyperbolic tree with three search terms

By selecting that entity, the tree expands an additional level and displays all entities related to it. The officer may choose to view the hyperbolic tree using filters given on the bottom of the screen.

These include filters on different entity types as well as a slider that allows the viewing of a given number of results (in Fig. 8 the officer has used the slider to view the top 15 results according to the weight of the relationships). The officer can view any section of the tree by moving it to the center.

• COPLINK Spatio Temporal Visualizer

The Spatio Temporal Visualizer (STV) takes COPLINK one step further by providing an interactive environment where analysts can load, save, and print police data in a dynamic fashion for exploration and dissemination. It overcomes some of the disadvantages of other existing crime visualization tools by viewing three perspectives on the same data. It show map (location) and temporal (time) data and allows the user to perform five types of tasks: Overview, Zoom, Filter, Relate, and Extract.

In this case, bank robberies for the last six years are displayed in the timeline, GIS and periodic views. From here, users may narrow focus through granularities and time bounds as well as geographic parameters.

Figure 9.

Views may be moved to provide better focus or because of user preference. Here, GIS view is centered and a geographic query is performed. The data set is narrowed to those selected by the user with corresponding updates in other tools. In the timeline view, points within the geo-search are emphasized, while other points are faded. The periodic view displays summary data on the selected points indicating June, April, November and December have higher incidence of bank robberies. The control panel allows for focus onto a specific period of time within the global time frame selected. Granularity (viewing in terms of days, weeks, months, years) and global time bounds may also be altered.

• COPLINK Criminal Network Analysis

The COPLINK CrimeNet Explorer is a system for exploring criminal networks in law enforcement and intelligence domain. It can help identify the central members, detect the groups, and extract the structure/organization in criminal networks based on criminal-justice data. The major technologies used are Social Network Analysis (SNA) methods (centrality measures and blockmodeling), clustering, concept space approach, and multidimensional scaling (MDS). It utilizes a network representation and facilitates six types of tasks: Overview, Zoom, Filter, Details-on-demand, Relate, and Extract.

Figure 10. A 57-member Gang network: Nodes represent individual criminals labeled by their names. Links represent relationships between criminals.

Applications

Intelligence Analysis and Knowledge Discovery

COPLINK Solution Suite has strong data analysis capabilities as well as strong data visualization capabilities. These capabilities make it as a useful tool for law enforcement units during their investigation. COPLINK has a lot of success stories in many states such as California, Arizona, Florida etc^[29].

Surveillance and Emergency Response

Infectious disease outbreaks, either naturally occurred or caused by biological terror attacks, pose a critical threat to public health and national security. Information systems and infectious disease informatics research are playing an increasingly important role in developing a comprehensive approach to prevent, detect, respond to, and manage infectious disease outbreaks. COPLINK Spatio Temporal Visualizer(STV) technology has also been applied to biosurveillance. It is applied to the BioPortal project developed in AI Lab at University of Arizona to help visualize biosurveillance data for disease surveillance purpose.

Case Study 2: i2 Analyst's Notebook

Introduction

Analyst's Notebook is a visual investigative analysis software developed by i2 Inc., which is the leading worldwide provider of visual investigative analysis software for law enforcement, intelligence, military and Fortune 500 organizations. Analyst’s Notebook provides the optimum environment for effective link and timeline analysis. The current version for Analyst's Notebook is version 6.

Analyst's Notebook is the de-facto standard software for visualisation and analysis adopted by more than 12,000 users in over 1,000 law-enforcement and commercial organisations world-wide^[30]. On April 3, 2006, The U.S. Federal Bureau of Investigation (FBI) expanded its use of i2® products under a new five-year licensing agreement worth an estimated $12 million^[31].

Overview

With Analyst's Notebook 6 users can conduct analysis to quickly reveal the relationships and patterns hidden within their data. The connections between people, organizations, accounts, phone records or any other element are then displayed in intuitive charts. The insights gained from both link and timeline analysis can be displayed in a single "hybrid" analytical chart.

Anayst's Notebook 6 provides a number of automatic chart rearrangement tools that reposition the elements on the chart to aid comprehension and provide more revealing visualization:

• Grouped Layout arranges the elements of the chart to emphasize the groupings present in the data.

Figure 11. Grouped Layout within Pictures

• Peacock Layout arranges the chart to emphasize the nodes and links that actually connect groups together, rather than the groups themselves.

Figure 12. Link Analysis in Peacock Layout

• Hierarchy Layout places the elements in a tree-like structure, particularly useful for showing organizational hierarchies. (See Figure 4. in Literature Review)

• Circular Layout positions the elements around the circumference of a circle to easily identify those chart elements with many links.

Analyst’s Notebook 6 also provides tools to perform a range of analytical tasks:

• Link Analysis visually links different entities, such as people and organisations, to show the relationships between them. (See Figure 1. in Literature Review)

• Network Analysis applies link analysis to large data sets such as telephone calls, bank account transactions, and Internet traffic records, to reveal paths, clusters or connected groups.

Figure 13. Network Analysis

• Timeline Analysis depicts events as they unfold over time, allowing you to understand cause and effect, identify patterns, and decide upon appropriate courses of action.You can also track entities over time, to identify their associations with an event or incident.

Figure 14. Timeline Analysis

• Transaction Pattern Analysis allows you to display a large series of transactions in chronological order. This helps to reveal repeating patterns of activity and can help to predict future behaviour.

Figure 15. Transaction Analysis

Applications

Intelligence Analysis and Knowledge Discovery

Intelligence analysis is performed by dedicated analysts working in law enforcement, government and military organizations. Intelligence analysts work to assemble and process all available information, whether it is from open sources such as the world wide web or closed sources such as intelligence records in internal databases.The challenge for intelligence analysts is assembling, grading, and analyzing vast amounts of data in a wide variety of formats.

Analyst's Notebook can assist intelligence analysts by providing a suite of tools that enables them to organize and analyze disparate data within a common platform. In every step of the intelligence analysis process, Analyst's Notebook offers software to help analysts quickly turn raw data into actionable intelligence.

Figure 16. Jemaah Islamiyah Terrorist Chart created by Analyst's Notebook

Terrorism Informatics and Countermeasures

Terrorists always rely on a series of loosely connected criminal activities to fund and carry out their objectives including: fraud, forgery, drug trafficking, money laundering, INS violations, murder and kidnapping. Analyst's Notebook provides a suite of tools that enable investigators to capture and organize data from a wide variety of sources, discover the hidden connections in disparate data sets and effectively communicate and share intelligence with cooperating agencies.

Figure 17. President Bush looks over a chart created with i2 software depicting Osama bin Laden's financial network during a tour of the Department of Treasury's Financial Crimes Enforcement Network in Vienna, Virginia, Wednesday, Nov. 7, 2001. (AP Photo/Doug Mills)

Cast Study 3: Starlight

Introduction

Starlight, built by Battelle at the Pacific Northwest National Laboratory (PNNL), supports analysis of dynamic and complex information streams that consist of structured and unstructured text documents, measurements, images, and video data. It is a 3-D visualization system of multimedia information and it is a forerunner of an emerging new class of information system.

Starlight was used by Department of Defense to battele Y2K in 1999^[32] and in 2005 it was also used to help to track bioterrorism in Seattle^[33].

Overview

Starlight ingests large quantities of data, required to be in Extensible Markup Language (XML) format, and generates stunning visualizations of structured values, conditioned with color and shape in a manner not that different from the symbology and thematics used in mapping. Starlight also supports natural language queries in a manner similar to Google or Yahoo^[34]. Starlight visualizes various types of information, including structured and unstructured text, geographic information, and digital imagery. These multimedia information are visualized in 3-D space altogether. Relational information are linked together.^[35].

Starlight has its own Information Model which attempts to effectively capture relationships among information objects. The Starlight Model is comprehensive, capable of accommodating a wide variety of relationship types, including discrete property (i.e., field/value pair) co-occurrences, free-text similarity, temporal relationships, parent-child associations, network relationships, and spatial (e.g., geospatial) relationships.

The following table is a summary of components of the Starlight Information Model (Data Representation).

Table 1. Components of the Starlight Information Model.

• General Similarity: A visualization workflow begins with all the data displayed as a data sphere. This shows a uniform view of data collections and their relationship by shape and color. See Figure 17.

Figure 18.

• Network: Normally, a network view could just represent the lines of coincidence between nodes on the network. Starlight can represent the relationships of coincidence, thematize the nature of those relationships and show the sub-relationships as well. Remember that all the attributes are available just like in a mapping program.As this is 3D, you can rotate the object to look at the other side. (See Figure 2. in Literature Review)

• Hierarchical: This image might represent a data visualization of your hard drive, with each cluster being a folder or directory and the file that they contain.In my case, the top could represent my root directory (C:\) and the bottom Pan would be either my temporary directory or more likely my email directory and all the old emails they contain. See Figure 19.

Figure 19.

• Mutlidimesional, Spatial, Temporal: The following figure shows a "Link Array" which contains Multidimesional, Spatial and Temporal data representation. The link array works similar to a prism map where the geography is pulled up the Z axis by a value.In this case (starting at the bottom), a data grid is pulled up by a value, tied to a data relationship with the image of the airplane, and then linked to two zoomed sections of a data grid prism map. Note that none of the data has to be from the same source, just related information. (See Figure 3. in Literature Review)

From the above figures, we can see that Starlight certainly expands the view of geography and perhaps offers an insight to skill sets that are in our future.Which is bringing together real, virtual and cyber geography into one 3- and 4-D workspace, performing better and more thorough analysis and ending up with far better results^[34].

Applications

Intelligence Analysis and Knowledge Discovery

Starlight can be used in Web Mapping. The Starlight Network View can be used to find and interpret interesting features in Web page hyperlink structures. When coupled with Starlight's Concept View text visualization capabilities, entirely new forms of Web exploration become possible.

Infrastructure Protection and Cyber Security

Starlight can be used in Network Security. Computer network security has become an increasingly pressing issue for many organizations. Starlight can enable network analysts to quickly achieve and maintain an in-depth understanding of network vulnerabilities and security status.

Terrorism Informatics and Countermeasures

Starlight can be used in National Security. Starlight's information integration capabilities make it uniquely well suited to analyzing the contents of multisource intelligence collections. With its integrated information extraction and geospatial analysis tools, Starlight users can quickly discover the "who, what, when, and where?" aspects of complex, dynamic situations.

Conclusions and Future Directions

In this chapter, we have provided a theoretical foundation for visualization. We have also provided a framework for visualization techniques, and evaluated some intelligence and security informatics tools within the framework. We have reviewed system design, as well as some evaluation techniques for ISI visualizations. We have seen various examples of tools for different ISI tasks, and examined three of these tools in greater detail. From our review, we have seen the importance of visualization tools and techniques for security and intelligence-related tasks. In the future, we believe there will be a strong focus on incorporating these and other emerging tools in the field of ISI.

References

1. ↑ Tegarden, D. P. (1999). Business information visualization Communications of the AIS 1(1es).
2. ↑ ^2.0 2.1 Bertin, J. (1983). Semiology of Graphics: Diagrams, Networks, Maps. University of Wisconsin Press.
3. ↑ Card, S. K., J. D. Mackinlay and B. Shneiderman, Eds. (1999). Readings in information visualization: using vision to think, Morgan Kaufmann Publishers Inc.
4. ↑ Nielson, G. M. (1991). Visualization in scientific and engineering computation. Computer 24(9): 58 - 66.
5. ↑ Robertson, G., S. K. Card and J. D. Mackinlay (1993). Information visualization using 3D interactive animation Communications of the ACM,.
6. ↑ ^6.0 6.1 Shneiderman, B. (1996). The eyes have it: a task by data type taxonomy for information visualizations.
7. ↑ Lamping, J., R. Rao and P. Pirolli (1995 ). A focus+context technique based on hyperbolic geometry for visualizing large hierarchies Proceedings of the SIGCHI conference on Human factors in computing systems, Denver, Colorado, United States ACM Press/Addison-Wesley Publishing Co.
8. ↑ Hsinchun Chen, A. L. H. R. R. S. B. R. S. (1998). Internet browsing and searching: User evaluations of category map and concept space techniques. Journal of the American Society for Information Science 49(7): 582-603.
9. ↑ ^9.0 9.1 Hearst, M. A. (1995). TileBars: visualization of term distribution information in full text information access. Conference on Human Factors in Computing Systems, Denver, Colorado, United States.
10. ↑ Donath, J. (2002). A SEMANTIC APPROACH TO VISUALIZING ONLINE CONVERSATIONS. Communications of the ACM 45(4): 45-49.
11. ↑ ^{11.0 11.1 11.2} Eick, S. C., J. L. Steffen and E. E. Sumner, Jr. (1992). Seesoft-a tool for visualizing line oriented software statistics. Software Engineering, IEEE Transactions on 18(11): 957-968.
12. ↑ Chen, C. and R. J. Paul (2001). Visualizing a knowledge domain's intellectual structure. Computer 34(3): 65-71.
13. ↑ ^{13.0 13.1 13.2} Ware, C. (2000). Information visualization: perception for design, Morgan Kaufmann Publishers Inc.
14. ↑ Collins, A. M. and E. F. Loftus (1975). A Spreading-Activation Theory of Semantic Processing. 82, 6, 407-28, Nov 75.
15. ↑ Yufik, Y. M. and T. B. Sheridan (1996). Virtual networks: new framework for operator modeling and interface optimization in complex supervisory control systems. Annual Reviews in Control 20: 179-195.
16. ↑ Tufte, E. (1990). Envisioning information. (Graphics Press).
17. ↑ Gershon, N. and S. G. Eick (1997). Information Visualization. Computer Graphics and Applications, IEEE 17(4): 29-31.
18. ↑ Mackinlay, J. (1986 ). Automating the design of graphical presentations of relational information ACM Trans. Graph. ACM Press 5 (2): 110-141
19. ↑ Salton, G., J. Allan and C. Buckley (1994). Automatic structuring and retrieval of large text files Communications of the ACM 37(2): 97-108.
20. ↑ ^20.0 20.1 A. Abbasi and H. Chen, "Visualizing Authorship for Identification", IEEE International Conference on Intelligence and Security Informatics, ISI 2006, San Diego, CA, USA, May 2006, pp. 60-71, Proceedings.
21. ↑ L. Deligiannidis, A. Sheth, and B. Aleman-Meza, "Semantic Analytics Visualization", IEEE International Conference on Intelligence and Security Informatics, ISI 2006, San Diego, CA, USA, May 2006, pp. 48-59, Proceedings.
22. ↑ T. Buennemeyer, J. Booker, A. Sabri and C. North, "Intelligence Analysis Using High Resolution Displays", IEEE International Conference on Intelligence and Security Informatics, ISI 2006, San Diego, CA, USA, May 2006, pp. 652-653, Proceedings.
23. ↑ P. Yan, D. Zeng, and H. Chen, "A Review of Public Health Syndromic Surveillance Systems", IEEE International Conference on Intelligence and Security Informatics, ISI 2006, San Diego, CA, USA, May 2006, pp. 249-260, Proceedings.
24. ↑ C. Yang, N. Liu, and M. Sageman, "Analyzing the Terroristic Social Network with Visualization Tools", IEEE International Conference on Intelligence and Security Informatics, ISI 2006, San Diego, CA, USA, May 2006, pp. 331-342, Proceedings.
25. ↑ E. Reid, and H. Chen, "Mapping the Contemporary Terrorism Research Domain", Int. J. Human-Computer Studies 65 (2007), pp. 42-56.
26. ↑ A. Khashman and A. Garad, " Intelligent Face Recognition Using Feature Averaging", IEEE International Conference on Intelligence and Security Informatics, ISI 2006, San Diego, CA, USA, May 2006, pp. 432-439, Proceedings.
27. ↑ S. Park and M. Trivedi, " Multi-perspective Video Analysis of Persons and Vehicles for Enhanced Situational Awareness", IEEE International Conference on Intelligence and Security Informatics, ISI 2006, San Diego, CA, USA, May 2006, pp. 440-451, Proceedings.
28. ↑ O. Kursun, K. Reynolds, and O. Favorov, " Matching Images of Never-Before-Seen Individuals in Large Law Enforcement Databases", IEEE International Conference on Intelligence and Security Informatics, ISI 2006, San Diego, CA, USA, May 2006, pp. 766-767, Proceedings.
29. ↑ ^29.0 29.1 "COPLINK Successful Stories" http://www.coplink.net/success.htm
30. ↑ http://www.i2inc.com/
31. ↑ "FBI Awards Contract to Deploy Agency-Wide Investigative Analysis Software "http://www.i2inc.com/Company/Press/040306.asp
32. ↑ "DoD Uses New Information Technology to Battle Y2K" http://www.defenselink.mil/news/Nov1999/n11011999_9911011.html
33. ↑ "Detectives track terrorists by Starlight" http://seattlepi.nwsource.com/national/251828_terror13.html
34. ↑ ^34.0 34.1 Starlight Overview and Interview with Battelle's Brian Kritzstein By Hal Reid , Senior Technical Editor, Directions Magazine & Private Consultant March 08, 2005 [1]
35. ↑ J.S. Risch, D.B. Rex, S.T. Dowson, T.B.Walters,R.A. May, B.D. Moon, The STARLIGHT Information Visualization System, Proceedings of IEEE International Conference on Information Visualization, IEEE, 1997