Current CERI Projects

Communication Analysis of Chat Data in the Air Operations Center

Communication among members of a team can be considered cognitive processing at the team level.  Therefore, the cognitive state of a team (e.g., situation awareness, shared understanding) can be assessed through an analysis of communication.  We have been working on ways to automate the analysis of team communications for real-time monitoring and assessment of team performance.   Toward that goal we have been exploiting the information that is available in patterns of turn taking or communication flow from person to person.  Our previous work in this area has largely focused on voice communications of three people in an Unmanned Aerial Vehicle context.  We have developed a tool kit for analyzing flow patterns called FAUCET.

In this project, CERI is working with Aptima to extend FAUCET to the context of the Dynamic Targeting Cell of the Air Operations Center.  Communications in this cell consists of a dozen or so individuals communicating primarily via chat.  These communications are embedded within the larger Air Operations Center comprising 100s of personnel.  The general goal of this research is to identify meaningful patterns of communication flow data that can ultimately serve as markers for real-time or predictive team interventions.  This work is sponsored by an STTR grant to Aptima by the Air Force Office of Scientific Research.

Click here for relevant publication on Team Communication Analysis


The CERTT Laboratory has had a long-standing relationship with the Air Force Office of Scientific Research (AFOSR).  AFOSR has most recently funded the CERTT Lab’s seventh and eight experiments.  AF7, which was completed in 2005, explored the nature of the acquisition and retention of team-level skills such as communication and coordination while also examining the effect of keeping teams intact or mixing individuals after a break of several weeks.  The findings that mixed teams showed improvement in process and performance after the break due to increased adaptability in the face of environmental and task perturbations became the basis of AF8.  AF8 explored three different training regimen: shared mental model, local optimal model, and perturbed.  It was hypothesized that by being exposed to increased adversity, perturbed teams would show better performance after a break of several weeks.  We have just completed data collection and data analyses are currently ongoing. 

Next Generation Pathfinder Architecture

Pathfinder, also known as KNOT (Knowledge Network Organizing Tool), was developed to aide researchers in deriving cognitive representations of domain knowledge.  Pathfinder has been long used by AFRL for assessing Distributed Mission Operations (DMO) training effectiveness.  The Pathfinder scaling algorithm transforms a subject’s judgment of the relatedness of pair-wise combinations of concepts in the domain into a network representation of the subject’s knowledge of concepts in the domain.  The relatedness of concept pairs is usually described as proximities or conversely distances.  The network is often called a Pathfinder network or PFNET, and consists of the items as nodes and a set of links.  The set of links is determined by patterns of proximities in the data and parameters of Pathfinder algorithms.  Developed in the early 1990s, existing Pathfinder tools are stand-alone DOS applications and are thus cumbersome to embed and integrate into complex and existing software environments.  There was a need to upgrade the applicability of Pathfinder tools by providing researchers with both stand-alone analysis tools, such as advanced Windows-based applications, and components for use as embedded objects within other psychological applications (e.g. DLL, DCOM, and/or ActiveX).  In this effort, CERI developed the next generation Pathfinder architecture.  The new architecture now defines Pathfinder to be much more than an improved algorithm; the next-generation tools will consist of a complete system architecture that encompasses data collection, network exchange, analysis, and network display.  Additionally, the scope of Pathfinder has been expanded.  As an example, the original Pathfinder primarily used textual proximities (distances) as relatedness data.  In the new architecture, our proximity data collection scheme uses more than semantics; it now has the capability to present participants with stimuli such as video, audio, text, documents, and images.   

Basic Research for Communication Analysis

Researchers working at CERI have a long-standing relationship with the Office of Naval Research (ONR).  ONR has funded three basic research grants to these researchers covering the topics of communication analysis and automation of communication analysis.  Previous efforts have included using Latent Semantic Analysis and low-level communication flow data to predict performance, identifying issues in face-to-face versus distributing teamwork using communication content and flow data, and measuring workload through communications.  Current partnerships between CERI researchers and ONR are centered on identification of dynamics that occur in communications that generalize across many work domains, using communication flow data.  One aspect of this communication research that is particularly novel is that communication flow data are domain- and language-independent.  The CERI researcher-ONR connection can capitalize on this feature in order to apply custom communication analysis methods to a wide range of cognitive engineering problems.


The current research objective is to identify valid and reliable measures of behavioral change, behavioral intent, motivations, perceptions, and attitudes under the assumption that the causal factors are based on psychological operations.  The Cognitive Engineering Research Institute (CERI) research team is currently conducting a comprehensive academic literature review of material related to the reliability and validity of measures associated with behavioral change. .  This research ultimately aims at  providing an understanding of the most valid and reliable methods for measurement of behavioral change with regard to psychological operations.

CERI Home PageMissionOrganizationScience & TechnologyMedia & PublicationsResearch CollaborationsFacilitiesSpecial ProgramsDonations & SponsorshipsCERI Message Board (will open new window)
CERI WorkshopsWomen in Science ProgramCERI Community OutreachCERI Brown Bag
Cognitive Engineering Research on Team Tasks Lab
PublicationsCERI News & ArticlesCERI PicturesCERI MoviesOther Documents
Current CERI ProjectsPast CERI Projects
StaffCERI Science DirectorCERI Organization ChartCERI Board of DirectorsCERI Institutional Review Board
CERI Home Page
About CERI CERI Contact Information Visiting CERI CERI Publications CERI Website Site Map CERI Employee Section
This page was last updated on 7/11/2007