43 results about "Email classification" patented technology

A method, and corresponding system, for identifying e-mail messages as being unwanted junk or spam. The method includes converting the outputs of a set of e-mail classification tools into a standardized format, such as a probability having a value between zero and one. The standardized outputs of the classification tools are then input to a voting mechanism which uses a voting algorithm based on fuzzy logic to combine the standardized outputs into a single classification result. The use of a fuzzy logic algorithm creates a more useful result as the classifier results are not merely averaged. In one embodiment, the single classification result is itself a probability that is provided to a spam classifier or comparator that functions to compare the single classification result to a spam threshold value and based on the comparison to classify the e-mail message as spam or not spam.

A system for intelligently sorting e-mail comprises a client which downloads e-mails from a server. Prior to presenting the e-mails to the user, the client sorts the e-mail into classifications based upon whether the e-mail is from a personal contact, i.e. someone that that the user knows, whether the e-mail is from a commercial vendor from whom the user has indicated that he or she wishes to accept commercial e-mail, or whether the e-mail is from an unknown source. The client presents the e-mails to the user in these classifications.

A system and method for establishing temporary electronic mail (e-mail) accounts is configured to operate with classified newspaper advertising. One of the aspects of the present invention is a file handling method that automates the process of establishing the e-mail accounts through interaction with a publisher's system for taking orders for classified advertisements. Typical newspaper classified advertising comprises two or three lines of text which provide only basic information about an item for sale. With the present invention, advertisers would be offered the option of creating a special e-mail account with a duration tied to the run of the advertisement. The e-mail accounts would be used to receive inquiries about the item for sale. The e-mail account could be also be configured to auto-respond to each incoming message with additional descriptive information about the item being offered for sale. Responders could learn more about the item, and thereby avoid the necessity of calling the seller. Sellers, in turn, can avoid the problem of repeated phone calls to request additional information, and may also remain anonymous while responding to multiple inquiries.

Methods, apparatuses, and computer-readable media for automatically disabling disposable e-mail addresses based on user actions. A method embodiment for disabling disposable e-mail addresses comprises: receiving (205) at a disposable e-mail address server (DEA server) (330) an e-mail directed to the DEA server (330) via a disposable e-mail address, wherein the disposable e-mail address is associated with an un-aliased e-mail address; adding (220) tracking information to the e-mail; directing the e-mail to a e-mail client; classifying (245) the e-mail as unsolicited e-mail; ascertaining (260) creation data associated with the disposable e-mail address; conveying the creation data for subsequent interpretation; seeking (270) authorization to disable the e-mail address; and upon gaining authorization, directing (275) the DEA server (330) to disable the disposable e-mail address.

In an electronic mail system, an apparatus and method for providing, in addition to the conventional “Inbox” directory in which all incoming e-mails are normally received, a plurality of appropriately labeled directories for containing e-mails suspected of being spam, grading the level of spaminess of the incoming e-mails and then moving or copying the incoming e-mails into one or more of the spam directories based upon the e-mails' respective levels of spaminess.

An electronic mail management system is operable to receive electronic mail message components from a user. Prior to sending of the email message, a plurality of predetermined categories for classifying the electronic message are presented to the user. The email system receives a user selection from the predetermined categories. A determination is made as to whether or not the email recipient is a member of a predetermined class of recipients. The electronic mail is then sent to the recipient. The user selection is embedded within the email message if the recipient is a member of the predetermined class of recipients. Finally, the email is stored in a manner dependent upon the user selection. Similarly, the email management server may also be operable to categorize received email messages. Further, email messages may be selected or programmed to automatically be deleted upon certain events.

The present invention generally relates to an improved system and method for providing email classification. Specifically, the present invention relates to an email classification system and method for analyzing the signature of an email for proper classification.

An information processing system including: a first importance level estimation unit that employs information for emails previously received by a user to classify an email to one of a plurality of first importance level categories; and a second importance level estimation unit that employs the information for the emails previously received by the user and employs a determination reference that differs from the one used by the first importance estimation unit, to determine importance level of emails belonging to each of the first importance level categories from one of a plurality of second importance level categories, which differ from the first importance level categories, the second importance level estimation unit designating the second importance level categories in correlation with the first importance level categories, respectively. Information for the second importance level categories obtained by the second importance level estimation unit is subjected to a predetermined process.

An e-mail categorization system can include a plurality of role templates, each role template being associated with an e-mail account and defining a plurality of categories for categorizing e-mail messages. An authorized account holder can establish the plurality of categories and can designate programmatic actions that are to be selectively executed by an e-mail receiving application upon the receipt of e-mail messages. The selection of actions can be based upon categories of the received e-mail messages. Role templates can be distributed to e-mail originators so that when e-mail originators direct e-mail messages to an account associated with a role template, the originators are prompted to select one of the categories defined by a corresponding role template.

After a client or potential client accesses the web-site of a host organization and generates an e-mail to the organization, an e-mail sorting and routing system parses the meta-tags appended to the message to appropriately sort and route the message. The meta-tags specify the language of the client, i.e., the language in which the web-site was communicating with the client when the message was generated. The meta-tags also specify the topic, e.g. a product or service, about which the message was written. This information may be input by the client or ascertained from the content or purpose of the web-page from which the e-mail was generated. The messages are then sorted, first by language and then by topic. As a result, each message is quickly routed to a person qualified as to both language and subject matter who can respond to the client.

The invention is selectively used by an Email Protection System that protects and email system from spam. This method is used if the Email Protection System classifies as spam an email connection or email message from an Email Sender. The invention is unlike other white list population methods or challenge / response systems in that it requires positive action by both the Email Sender and the Intended Email Recipient to add the Email Sender to a White List used by the Email Protection System. In accordance with the invention, the Email Protection System notifies the Email Sender of rejected email. The Email Sender uses a Request Form to request to be added to a White List. The Email Recipient confirms or denies the request. If the Email Recipient confirms, the Email Sender is added to the White List.

The invention discloses a multilevel Email classification method based on Email content. For the problem that imbalance of samples has significant impact on an Email classification effect, a mode of establishing three-level classifiers is employed. According to the mode, a naive bayes algorithm, a support vector machine and a C4.5 algorithm combined voting mode is employed in establishment of a first level classifier; a random forest algorithm is employed in the establishment of a second level classifier; and a liblinear algorithm is employed in the establishment of a third level classifier. According to the method, for design of the classifiers, a multilevel classification mode is employed; levels of the algorithms are increased; through layer by layer filtering, under the condition of ensuring a recall rate, the accuracy of the classifiers is improved gradually; the balance of positive and negative samples is increased continuously; basic balance of the Email data is realized at the last level; the final classification effect is less influenced by the imbalance of the samples; and the relatively good effect is realized in Email classification.

A class level of an email is determined responsive to defined rules for email classification. Client access to the email is then controlled responsive to the class level. One of a plurality of client authentication modes may be selected for authentication of the client responsive to the class level. The class level of the email may be determined responsive to an operational use of the email, a known keyword being present in content of the email, a source address of the email, a number of destination addresses of the email, an addressed resource of a user selectable link that is contained in a body of the email, characteristics of earlier email that the client deleted without reading, and / or various other rules.

An embodiment of the invention discloses an email classification method and an email classification device. The method includes performing word segmentation on texts in emails and acquiring word segmentation results; using TF-IDF (term frequency-inverse document frequency) technology to sequence the word segmentation results; acquiring similarity between each two emails; clustering the emails according to the similarity and acquiring clustering results; displaying the clustering results. By the method and the device, when a user do not want perform any operations on the emails, the emails can be classified intelligently according to content of the emails, so that work of the user can be reduced; when the user intervenes classification results, priority sequencing can be performed according to rules set by the user, so that operation flexibility is improved, and experience and convenience of the user of the emails are improved.

The invention discloses a word-sentence-level short text classification method based on deep learning, and belongs to the technical field of natural language processing. According to the method, wordfeatures and sentence features are combined based on a word vector technology to express complex text features. By performing convolution pooling on the word vectors in the single sentence through a plurality of convolution kernels of the convolutional neural network, all the feature maps are connected to obtain sentence vectors, and the sentence vectors are inputted into the long-term and short-term memory network according to a time sequence to perform context association so as to better express text contents. To-be-classified short text data is subjected to sentence segmentation, word segmentation, stop word removal, word vector conversion and the like and then input into a word-sentence level convolutional recurrent neural network for training, and finally, a short text classificationmodel can be obtained and a short text classification task is completed. The method has good performance in the aspects of tested Chinese spam e-mail classification and news text classification.

A sender-driven framework for enhancing the categorization of emails, wherein the sender is able to create or define rules that refine or augment those at the receiver's end. Particularly, the sender is enabled, in accordance with at least one embodiment of the present invention, to add information relating to the context of an email, which can help streamline categorization at the receiver's end.

The invention discloses an e-mail sorted storing method and device. The e-mail sorted storing method comprises the first step of receiving new mails and acquiring the e-mail addresses of addressers of the new mails, the second step of determining group names of contact persons, corresponding to the e-mail addresses of the addressers, in a preset address book, and the third step of storing the new mails into folders named as groups which the contact persons belong to. By means of the method, the groups which the addressers belong to are determined according to the groups of the contact persons in the address book; when the mails of the corresponding contact persons are searched for, a user just needs to search for the mails in the folders corresponding to the contact persons and named as the groups of the contact persons; accordingly, the search range is shrunk, the problem that mails of special contact persons are hard to find quickly in a large number of e-mails is solved, the group names of the contact persons of the mails are made to be the same as the group names of the contact persons in the address book, and therefore a mobile terminal has higher consistency in contact person storage.

A method for classifying email includes receiving an email. Several recipient email accounts of the email are extracted. Several email feature values of the email are generated according to the recipient email accounts. A classification algorithm is utilized to classify the email as an official email or a private email according to the email feature values of the email.

A method and system method for e-mail management of e-mails having embedded classification metadata. A query from an end user to access an e-mail account by an e-mail client is received. It is then determined whether the query has come from an insecure e-mail client, such as a web e-mail client. Access rules for defining classification access restrictions for the e-mail client to access e-mail in the e-mail account are retrieved when the e-mail client is determined to be insecure. The e-mail query can then be modified before sending to an e-mail server storing the email account. The query is modified based on the retrieved access rules to exclude retrieval of e-mails based upon the e-mail classification metadata.