Publications -> Journal Papers

Neurocomputational Modeling of Prospective Memory


Authors: J. Hou and X. Tao
Title: Neurocomputational Modeling of Prospective Memory
Abstract: Forgetting is in common in daily life, and 50-80% everyday’s forgetting is due to prospective memory (ProM) failures, which have significant impacts on our life. More seriously, some of these memory lapses can bring fatal consequences such as forgetting a sleeping infant in the back seat of a car. People tend to use various techniques to improve their prospective memory performance. However, people with ProM difficulties (e.g., elderly) are often involved in many group tasks. The group of elderly participating in a group task may interact with each other before the date of the event and thus, a group member might be reminded by other group members of the group task. This thesis proposes a computational approach for determining the appropriate number of reminders and reminding method. The problem of determining the optimal reminding schedule is very complex due to many interdependent factors and uncertainty, it is thus difficult to build an integrated framework in which all the interdependent factors are concurrently optimized. Rather than modeling all the interdependent factors explicitly and then determine the optimal reminding schedule by an complicated combinatorial optimization, we try to connect all the inter-dependent factors indirectly and design novel heuristics to well approximate the optimal reminding schedule. In our approach, the reminding model will determine a reasonable number of reminders for a ProM task based on the predicted performance of the task. Guided by a reminder schedule function, the proposed model is also capable of generating an effective reminder schedule automatically. Moreover, the reminding model is able to make context-aware decisions regarding the reminding method. To evaluate the proposed reminding model, we conducted a preliminary user study and the participants felt that the reminders generated according to the reminding model are appropriate in terms of their number, schedule and reminding methods. The results also support that our approach provides a better overall experience and reminds more effectively than its control version.
Keywords: Prospective memory; Group task reminding; Fuzzy cognitive map
Journal Name: International Journal of Information Technology, vol. 23, no. 2
Publisher: Singapore Computer Society
Year: 2017
Accepted PDF File: Neurocomputational_Modeling_of_Prospective_Memory_accepted.pdf
Permanent Link: http://www.intjit.org/journal/download/down.php?file=/23/2/232_1.pdf
Reference: J. Hou and X. Tao, “Neurocomputational modeling of prospective memory,” International Journal of Information Technology, vol. 23, no. 2, pp. 1–19, 2017.
bibtex: 
@article{LILY-j41,
   author	= {Hou, Jinghua and Tao, Xuehong}, 
   title	= {Neurocomputational Modeling of Prospective Memory}, 
   journal	= {International Journal of Information Technology}, 
   year		= {2017},  
   volume	= {23}, 
   number	= {2}, 
   pages	= {1-19},
   publisher 	= {Singapore Computer Society}, 
}