Actors: a model of concurrent computation in distributed systems
Actors: a model of concurrent computation in distributed systems
Pricing via Processing or Combatting Junk Mail
CRYPTO '92 Proceedings of the 12th Annual International Cryptology Conference on Advances in Cryptology
Using Character Recognition and Segmentation to Tell Computer from Humans
ICDAR '03 Proceedings of the Seventh International Conference on Document Analysis and Recognition - Volume 1
EC '04 Proceedings of the 5th ACM conference on Electronic commerce
IBM Systems Journal
Zmail: Zero-Sum Free Market Control of Spam
ICDCSW '05 Proceedings of the Fourth International Workshop on Assurance in Distributed Systems and Networks (ADSN) (ICDCSW'05) - Volume 01
Cyberorgs: a model for resource bounded complex agents
Cyberorgs: a model for resource bounded complex agents
An intelligent spam filtering system based on fuzzy clustering
FSKD'09 Proceedings of the 6th international conference on Fuzzy systems and knowledge discovery - Volume 7
A scalable intelligent non-content-based spam-filtering framework
Expert Systems with Applications: An International Journal
Supporting many-to-many communication
Proceedings of the 2013 workshop on Programming based on actors, agents, and decentralized control
| Hi-index | 0.26 |
Not only does Spam put strain on computational and network resources, it also drains the attention resource of users. Existing approaches to combat spam make decisions on per sender or per message basis. Where the former demands explicit intervention by users, the latter suffers from false negatives and positives. We take an owned-resource view of email spam. An email recipient's attention space as well as computational and network resources devoted to handling email are viewed as precious owned resources. Anyone interested in using these resources - by sending a message which must be delivered and stored on a server and subsequently viewed in a list of received messages, and possibly read - must purchase the resource in the form of sending rights. These rights may be to resources owned by individuals and groups, and may also be owned by individuals or group. Our approach is based on the CyberOrgs model for encapsulating distributed owned resources for multi-agent computations. The model uses a small set of primitive operations to enable multi-agent computations to engage in a rich set of interactions in a market of owned resources. We have developed key mechanisms to enable message senders and recipients to negotiate contracts for mailbox access, as well as to construct policies leading to implicit and automatic negotiations for typical cases. A prototype implementation is described, and examples of policies are presented. Experimental results show that system performance is policy-dependent: as long as policies are carefully designed, negotiation overhead is minimal.