VRA Reporter for Field Monitoring and Analysis

For nearly two decades the VRA Reporter has been used by NGOs and IGOs as a field data monitoring and management tool supporting corporate security, staff safety, child protection and country stability around the world. The Reporter currently operates on a third-generation platform that is browser and device independent.

The Reporter offers customizable templates for incident and situation reporting as well as an optional text message feature for use when Internet access is not available. The Reporter supports data links across the templates and client customization of aggregation, analysis and visualization.

The Reporter includes a configurable dashboard with full search, filter and sort functionality to facilitate real-time understanding at a glance of the field data. The dashboard includes a selection of graphs and maps as well as a list view grid that dynamically linked to each user’s authorized view of the data.

The Reporter server operates on a single or two-tier server, and it can be hosted by VRA or the client, both backed by VRA’s optional reliable remote maintenance and support program. The Reporter server can be run with a secure (SSL) connection with a robust security model that supports differentiation among user roles, geographic locations and organizational memberships. User authentication is configurable as a stand-alone system, or it can be integrated with a client’s domain controller or with third parties such as Facebook or Google.

The Reporter supports multiple languages and can be integrated with collaboration, risk assessment and structural context modules. The following features are standard in the Reporter Module: Access for logging into the application & customer resource links Dashboard for viewing status and trends at a glance Incidents for managing Incident Reports (IncReps) Situations for managing Situation Reports (SitReps) Analysis for analyzing aggregated sets of reports Field Notes for managing SMS messages Preferences for managing user profiles & selecting preferences Administration for managing the system, content and users

VRA Assessor

The Assessor is VRA’s most recent product designed and developed to support Security Risk Assessments or SRAs. This interactive tool guides analysists through a series of steps to identify the threats to a given entity or activity, to describe their context, to select treatments to mitigate the threats, and their anticipated risk reduction in terms of impact and likelihood, and to calculate the residual risk after the treatments have been applied.

Each of the threat and treatment assessment calculations are explicit and transparent to facilitate review and comparative risk evaluation of the mitigation strategies. Reports are produced by each step in the risk assessment process:

Activity / Entity Description, via text narratives and annotations Context Risk Rating, with customizable categories and weights Threat Identification, from client-specified or default lists Raw Risk Assessment, rated on impact & likelihood Treatment Identification, from client-specified or default lists Anticipated Risk Reduction, rated on impact & likelihood Automated Residual Risk Calculations presented in A Security Risk Assessment, with printer friendly reports

The Assessor presents aggregated results for the context rating, threats and raw risk, treatments and anticipated risk reduction and the calculated residual risk. These aggregated results are presented as bottom line values for the SRA as a whole as well as discrete assessments for every threat-treatment pair to facilitate diagnosis of threat vulnerabilities and treatment mitigations.

VRA is currently developing an extension to the Assessor tool that will support tracking of the implementation of the SRA treatments. This extension will include the identification of responsible parties, a time line, and periodic evaluations. This extension will provide an evidence-based audit trail of an organizations risk reduction strategy.

VRA Advisor

The VRA Advisor is an automated sentiment extractor that assesses various dimensions of affect presented in unstructured text. Such sentiment is typically associated within various frames of named entities (e.g. countries, corporations, markets, or individuals). These sentiments are assessed within domains of interest such the economy, politics, and armed conflict. Refined and additional domains may be specified by users.

The Advisor calculates scores for each entity within the desired domains to reveal the strength of sentiment associated with each. The sentiment scores thus reveal the sentiment as expressed in the processed text toward the entities. The polarity of relevant sentiment extractions is scaled from negative to neutral / mixed, to positive, and includes a measure of magnitude for each.

Expressed sentiments associated with entities referenced in the processed text reveals the extent to which the entities are viewed or considered similarly, both favorably and unfavorably. Shared sentiment within a domain illuminates clusters of similar affect relevant to that domain. The resultant entity-sentiment profiles across domains yield an evidenced-based view of expressed views associated with each entity.

The Advisor’s default for the frame is countries, but the entities are user-specifiable. VRA uses three default macro-level default domains, politics, economy, and armed conflict (referred to as military). VRA has experimented with additional domains, for example, consumer demands and the environment. The former domain focuses on the expressed need for and consumption of goods & resources and the latter extension focuses on references to environmental pressures, pollution, natural disasters & accidents.

More generally, sentiment associated with an entity can be extracted from a text and assigned attributes along any number of dimensions using various parameters. The default dimensions include polarity, magnitude and scope. However, other dimensions of affect may also be implemented as required by a customer’s specific needs.

In sum, the Advisor offers a flexible tool for users to acquire information on the sentiment on a range of entities across a range of domains and with a range of measures. This approach to identifying relevant documents in the VRA Advisor is explicitly parsimonious; the design utilizes a relatively small number of terms that reflect higher level concepts rather than an exhaustive dictionary of discrete words. The Advisor has been successfully used to assess market sentiment over twenty-five years of news reports to determine the correlations between various domains and weekly market closing prices.