Market Driven Innovations and Scaling up of Twitris

From Knoesis wiki
Jump to: navigation, search

This PFI: AIR Technology Translation project focuses on translating Twitris’ collective social media intelligence technology to capabilities well beyond current state-of-the-art social media monitoring and analysis tools. The Twitris platform is important because it can provide collective exploitation of real-time social media streams, and a variety of relevant knowledge, to significantly improve decision-making and support timely actions in various domains of economic, human, and social development. Twitris’ unique features include real-time semantic analysis of social media content along spatio-temporal-thematic, people-content-network, and sentiment-emotion-intent dimensions. These features result in deeper, contextually-relevant analysis and actionable insights when compared to the leading competing technology in this market space. This project will result in a scale-up of Twitris.

This project addresses several technology gaps as it transitions Twitris from a research prototype to a scaled-up technology capable of supporting commercial applications. Consequently, three areas of research and technology enhancement will be conducted: 1) enhancing the functionalities of Twitris with a broad range of location-specific processing that requires addressing the challenge of scarcity of spatial metadata on Twitter, 2) semantics-enhanced filtering and improved user experience for automatic and semi-automatic filtering of tweets, which requires addressing challenges such as content ambiguity and information overload, and 3) scalable architecture supporting domain-specific, knowledge-enabled modules to handle high volume, variety and velocity of data.

In addition, the project will also provide a unique education and training platform for students and recent graduates to prepare them for careers involving entrepreneurship and business and economic development, and careers in startups. Specifically, the project (a) bridges basic research with technology development and intellectual property development that can lead to successful commercialization and (b) involves close collaboration with successful entrepreneurs, business partners, and customers. It will also undertake structured educational activities involving five technical and business courses, while continuing to foster much-needed diversity in high-tech fields and computer science. This project engages several business partners in strategically-important markets to carry out trials involving their customers in an effort to evaluate the efficacy and benefits of research and technology enhancements involved in this scale-up.


People

Principal Investigators: Prof. Amit P. Sheth
Collaborators: Jeremy Brunn, Pavan Kapanipathi, Alan Smith

Funding

Nsf.jpg

Social Media

Follow us on Twitter


Related Projects Using Twitris

NSF SoCS: Social Media Enhanced Organizational Sensemaking in Emergency Response

NIH eDrugTrends: Social Media Analysis to Monitor Cannabis and Synthetic Cannabinoid Use

Harassment: Context-aware Online Harassment Detection on Social Media

Project Safe Neighborhood (PSN): Westwood Partnership to Prevent Juvenile Repeat Offenders

Hazards SEES: Social and Physical Sensing Enabled Decision Support

Depression: Modeling Social Behavior for Healthcare Utilization in Depression

Twitris is also being used for graduate courses in Computer Science, Internet Marketing, and Management Sciences.

This project is a follow-on to

I-Corps: Towards Commercialization of Twitris — a system for collective intelligence: (NSF IIP-1343041). Outcome summary video

References

Twitris

News/Media

WSU Lab Works to Mine Social Media Posts, Dayton Daily News, July 13, 2016. [PDF]
[Discusses commercialization effort for Twitris developed at Kno.e.sis by Dayton startup Cognovi Labs, and quotes Prof. Sheth.]

The Twitris sentiment analysis tool by Cognovi Labs predicted the Brexit hours earlier than polls, TechCrunch, June 29, 2016. [PDF]
[Hours before the EU referendum votes closed and well before results were declared, Prof. Sheth’s analysis of Twitter data predicted #Brexit - votes for leave outpacing remain. Analysis was done using a campaign set up by Cognovi Labs that is powered by Twitris technology.]

Contact

Contact Prof. Amit P. Sheth for more details