Visualizing Host-Nation Sentiment at the Tactical Edge
Abstract
Measures of sentiment can help refine a Warfighter's knowledge and understanding of an unfamiliar operational environment, for example, the sentiment of civilians and insurgents to the Army, its operations, or its adversaries. In addition, observing changes in expressed sentiment about topics over time may provide baselines from which to detect if and when important shifts in attitude occur triggering further investigation. The paper discusses new visualization techniques for displaying sentiment offered by social media analysis platforms currently under development by the Army. The sentiment analysis capabilities of these platforms were tested within the realistic military conditions of the C4ISR OTM field exercise using an operational scenario composed of diversified data sources including social media content. A human subjects experiment is planned for more fine-grained testing to better understand the advantages offered by different visualization approaches to sentiment analysis. shift is under way in recent military operations from standard, kinetic warfare to stability, security, transition and reconstruction (SSTR) operations. There is growing recognition that new information sources and text processing techniques are needed by the intelligence community to understand the socio-cultural landscape of an area of operations. To meet this emerging need, OSD and ARL are making significant investments in purposefully funding and managing the development of SM analysis platforms with sentiment analysis and topic trending capabilities. Two SM analysis platforms currently under development, MiNPAC and MIST, extract and visually present social sentiment themes from large datasets. MiNPAC provides measures of author based sentiment how positive/negative are their communications to the public; while MIST provides community measures of sentiment how community members feel about their leaders in reference to topics of concern
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 2014
- Accession Number
- ADA606810
Entities
People
- Daniel N. Cassenti
- Elizabeth K. Bowman
- Heather E. Roy
- Sue E. Kase
Organizations
- United States Army Research Laboratory