| 摘要 |
An automatic message classification system includes a set of stored messages. Each message contains address identifiers. An incoming message processor delivers incoming messages to the system, each containing incoming address identifiers. A list of address identifiers is created by retrieving each address identifier of each message within the set of stored messages and adding the address identifier to the list if it does not already exist in the list. A central processor calculates a predictive address value for each address identifier in the list of address identifiers by dividing one by the number of occurrences of the address identifier in all messages within the set of stored messages. A central processor of the system compares the incoming address identifiers with the address identifiers of each message located within the set of stored messages to generate a predictive message value for each message. The predictive message value is the sum of the predictive address values for every incoming address identifier that also exists in the address identifiers of each message. The system classifies the incoming message with the message having the highest predictive message value.
|
