Intelligent Knowledge Distribution: Constrained-Action POMDPs for Resource-Aware Multiagent Communication

Information Distribution in Multi-Robot Systems: Generic, Utility-Aware Optimization Middleware

“…For instance, a method adjusted for distributed photogrammetry might not work well for a multi-robot search and rescue mission. Nevertheless, state of the art publications addressing the problem of information distribution are usually ignoring this aspect and evaluating the solutions using specific applications (Marcotte et al, 2019;Best et al, 2019;Unhelkar and Shah, 2016;Williamson et al, 2009;Fowler et al, 2020;Roth et al, 2006). Hence, in Section 4.1 we propose a novel evaluation methodology that presents the performance of our solution in different scenarios.…”

“…Usually, it is either taken for granted or the main focus is on maintaining the connectivity (Amigoni et al, 2019). Only recently, the problem of information distribution in MRS began to receive attention (Marcotte et al, 2019;Best et al, 2019;Fowler et al, 2020;Anderson and Hollinger, 2021).…”

“…They communicate the most valuable (i.e., providing the highest utility) pieces of information, assuming this would positively impact the mission performance. A common way to define such utilities is given by informationtheoretic measures such as KL-divergence (Fowler et al, 2020;Williamson et al, 2009). However, such approaches imply that the application logic takes into account information uncertainty.…”

“…This approach allows researchers to avoid the question of what to communicate and focus only on the question of when to communicate. Although in many cases it is fair to assume that all information could be transmitted in a single message, it is not applicable in a generic setting (Fowler et al, 2020). As an example, suppose the state of an agent consists of its position, battery level, and camera images.…”

Intelligent Recommendation Model of Contemporary Pop Music Based on Knowledge Map

“…In recent years, although the recommendation algorithms [ 5 ] in the fields of movies, news, and books have emerged one after another, there are not many recommendation systems in the field of music, mainly because the data in this field are often not made public, and unlike the ratings we usually use, there is basically no public data set that mentions the user's ratings [ 6 ]. Intelligent music recommendation system is an effective way to solve “information overload.” The recommendation system mines users' potential interest preferences according to their historical behaviors, calculates the similarity between users and resources, and recommends music resources that users may be interested in [ 7 ].…”