Start Date
11-8-2016
Description
The paper addresses the issue of including textual and network information into predictive modelling. Related work suggest models for both analytic fields separately. Considering a “Big Data World” where unstructured information and the often times neglected network information, both pose analytic challenges we aim at integrating both in a predictive model. In addition we propose a novel approach to capture dynamic network changes in network regularization of prediction models. While the model states a general approach the paper focuses on evaluation in the well-known stock market prediction challenge in order to make the research more comparable to related scientific work.
Recommended Citation
Heinrich, Kai, "The Missing Link – Predictive Models based on Textual and Dynamic Network Data" (2016). AMCIS 2016 Proceedings. 27.
https://aisel.aisnet.org/amcis2016/Decision/Presentations/27
The Missing Link – Predictive Models based on Textual and Dynamic Network Data
The paper addresses the issue of including textual and network information into predictive modelling. Related work suggest models for both analytic fields separately. Considering a “Big Data World” where unstructured information and the often times neglected network information, both pose analytic challenges we aim at integrating both in a predictive model. In addition we propose a novel approach to capture dynamic network changes in network regularization of prediction models. While the model states a general approach the paper focuses on evaluation in the well-known stock market prediction challenge in order to make the research more comparable to related scientific work.