Abstract

A key analytical task across many domains is model building and exploration for predictive analysis. Data is collected, parsed and analyzed for relationships, and features are selected and mapped to estimate the response of a system under exploration. As social media data has grown more abundant, data can be captured that may potentially represent behavioral patterns in society. In turn, this unstructured social media data can be parsed and integrated as a key factor for predictive intelligence. In this paper, we present a framework for the development of predictive models utilizing social media data. We combine feature selection mechanisms, similarity comparisons and model cross-validation through a variety of interactive visualizations to support analysts in model building and prediction. In order to explore how predictions might be performed in such a framework, we present results from a user study focusing on social media data as a predictor for movie box-office success.

Original languageEnglish (US)
Title of host publication2014 IEEE Conference on Visual Analytics Science and Technology, VAST 2014 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages193-202
Number of pages10
ISBN (Print)9781479962273
DOIs
StatePublished - Feb 13 2015
Event2014 IEEE Conference on Visual Analytics Science and Technology, VAST 2014 - Paris, France
Duration: Oct 9 2014Oct 14 2014

Other

Other2014 IEEE Conference on Visual Analytics Science and Technology, VAST 2014
CountryFrance
CityParis
Period10/9/1410/14/14

Fingerprint

Feature extraction
Visualization
Predictive analytics

Keywords

  • Feature Selection
  • Predictive Analytics
  • Social Media

ASJC Scopus subject areas

  • Computer Vision and Pattern Recognition
  • Computer Science Applications

Cite this

Lu, Y., Kruger, R., Thom, D., Wang, F., Koch, S., Ertl, T., & Maciejewski, R. (2015). Integrating predictive analytics and social media. In 2014 IEEE Conference on Visual Analytics Science and Technology, VAST 2014 - Proceedings (pp. 193-202). [7042495] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/VAST.2014.7042495

Integrating predictive analytics and social media. / Lu, Yafeng; Kruger, Robert; Thom, Dennis; Wang, Feng; Koch, Steffen; Ertl, Thomas; Maciejewski, Ross.

2014 IEEE Conference on Visual Analytics Science and Technology, VAST 2014 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2015. p. 193-202 7042495.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Lu, Y, Kruger, R, Thom, D, Wang, F, Koch, S, Ertl, T & Maciejewski, R 2015, Integrating predictive analytics and social media. in 2014 IEEE Conference on Visual Analytics Science and Technology, VAST 2014 - Proceedings., 7042495, Institute of Electrical and Electronics Engineers Inc., pp. 193-202, 2014 IEEE Conference on Visual Analytics Science and Technology, VAST 2014, Paris, France, 10/9/14. https://doi.org/10.1109/VAST.2014.7042495
Lu Y, Kruger R, Thom D, Wang F, Koch S, Ertl T et al. Integrating predictive analytics and social media. In 2014 IEEE Conference on Visual Analytics Science and Technology, VAST 2014 - Proceedings. Institute of Electrical and Electronics Engineers Inc. 2015. p. 193-202. 7042495 https://doi.org/10.1109/VAST.2014.7042495
Lu, Yafeng ; Kruger, Robert ; Thom, Dennis ; Wang, Feng ; Koch, Steffen ; Ertl, Thomas ; Maciejewski, Ross. / Integrating predictive analytics and social media. 2014 IEEE Conference on Visual Analytics Science and Technology, VAST 2014 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2015. pp. 193-202
@inproceedings{4c371842223646bcb4e066228c9ffa3f,
title = "Integrating predictive analytics and social media",
abstract = "A key analytical task across many domains is model building and exploration for predictive analysis. Data is collected, parsed and analyzed for relationships, and features are selected and mapped to estimate the response of a system under exploration. As social media data has grown more abundant, data can be captured that may potentially represent behavioral patterns in society. In turn, this unstructured social media data can be parsed and integrated as a key factor for predictive intelligence. In this paper, we present a framework for the development of predictive models utilizing social media data. We combine feature selection mechanisms, similarity comparisons and model cross-validation through a variety of interactive visualizations to support analysts in model building and prediction. In order to explore how predictions might be performed in such a framework, we present results from a user study focusing on social media data as a predictor for movie box-office success.",
keywords = "Feature Selection, Predictive Analytics, Social Media",
author = "Yafeng Lu and Robert Kruger and Dennis Thom and Feng Wang and Steffen Koch and Thomas Ertl and Ross Maciejewski",
year = "2015",
month = "2",
day = "13",
doi = "10.1109/VAST.2014.7042495",
language = "English (US)",
isbn = "9781479962273",
pages = "193--202",
booktitle = "2014 IEEE Conference on Visual Analytics Science and Technology, VAST 2014 - Proceedings",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - GEN

T1 - Integrating predictive analytics and social media

AU - Lu, Yafeng

AU - Kruger, Robert

AU - Thom, Dennis

AU - Wang, Feng

AU - Koch, Steffen

AU - Ertl, Thomas

AU - Maciejewski, Ross

PY - 2015/2/13

Y1 - 2015/2/13

N2 - A key analytical task across many domains is model building and exploration for predictive analysis. Data is collected, parsed and analyzed for relationships, and features are selected and mapped to estimate the response of a system under exploration. As social media data has grown more abundant, data can be captured that may potentially represent behavioral patterns in society. In turn, this unstructured social media data can be parsed and integrated as a key factor for predictive intelligence. In this paper, we present a framework for the development of predictive models utilizing social media data. We combine feature selection mechanisms, similarity comparisons and model cross-validation through a variety of interactive visualizations to support analysts in model building and prediction. In order to explore how predictions might be performed in such a framework, we present results from a user study focusing on social media data as a predictor for movie box-office success.

AB - A key analytical task across many domains is model building and exploration for predictive analysis. Data is collected, parsed and analyzed for relationships, and features are selected and mapped to estimate the response of a system under exploration. As social media data has grown more abundant, data can be captured that may potentially represent behavioral patterns in society. In turn, this unstructured social media data can be parsed and integrated as a key factor for predictive intelligence. In this paper, we present a framework for the development of predictive models utilizing social media data. We combine feature selection mechanisms, similarity comparisons and model cross-validation through a variety of interactive visualizations to support analysts in model building and prediction. In order to explore how predictions might be performed in such a framework, we present results from a user study focusing on social media data as a predictor for movie box-office success.

KW - Feature Selection

KW - Predictive Analytics

KW - Social Media

UR - http://www.scopus.com/inward/record.url?scp=84929494447&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84929494447&partnerID=8YFLogxK

U2 - 10.1109/VAST.2014.7042495

DO - 10.1109/VAST.2014.7042495

M3 - Conference contribution

SN - 9781479962273

SP - 193

EP - 202

BT - 2014 IEEE Conference on Visual Analytics Science and Technology, VAST 2014 - Proceedings

PB - Institute of Electrical and Electronics Engineers Inc.

ER -