Structure-based multi-intent email classification
AssigneeAccenture Global Solutions Limited
InventorsAnshuma Chandak, Abhishek Mukherji, Emmanuel Munguia Tapia
Life of this patent
Estimating the term…
Abstract
In some examples, structure-based multi-intent email classification may include receiving an email thread that includes a plurality of emails including an email that is to be classified, and identifying, for the email thread, a process associated with the email thread, Based on the process associated with the email thread and for each sentence of the email that includes a plurality of sentences, a corresponding classifier may be determined from a plurality of classifiers, and applied to each sentence of the email to generate classified sentences. At least one entity may be extracted from each of the classified sentences of the email. Based on the at least one extracted entity, the email may be classified.
Claims17
A structure-based multi-intent email classification apparatus comprising:
- an email thread analyzer, executed by at least one hardware processor, to
- receive an email thread that includes a plurality of emails including an email that is to be classified, and
- identify, for the email thread, a process associated with the email thread, wherein the process defines a logical path of contents of the email thread;
- a sentence classifier, executed by the at least one hardware processor, to
- select one or more classifiers of a plurality of classifiers based on transition probabilities of occurrences of the process that are affected by the reception of the email to be classified;
- determine, based on the process associated with the email thread and for each sentence of the email that includes a plurality of sentences, a corresponding classifier from the one or more classifiers that include at least two of a question, a statement, a command, or a request, and
- apply, to each sentence of the email, the corresponding classifier to classify each sentence of the email to generate classified sentences;
- an information extractor, executed by the at least one hardware processor, to
- extract, from each of the classified sentences of the email, at least one entity;
- and
- an email classifier, executed by the at least one hardware processor, to
- classify, based on the at least one extracted entity, the email.
A method for structure-based multi-intent email classification, the method comprising:
- receiving, by at least one hardware processor, an email thread that includes a plurality of emails including an email that is to be classified, wherein the email thread pertains to a process, that defines a logical path of contents of the email thread;
- identifying, by the at least one hardware processor, for the email thread, the process associated with the email thread;
- selecting, by the at least one hardware processor, one or more classifiers of a plurality of classifiers based on transition probabilities of occurrences of the process that are affected by the reception of the email to be classified;
- determining, by the at least one hardware processor, for each sentence of the email that includes a plurality of sentences, a corresponding classifier from the one or more classifiers that include at least two of a question, a statement, a command, or a request;
- applying, by the at least one hardware processor, to each sentence of the email, the corresponding classifier to classify each sentence of the email to generate classified sentences;
- extracting, by the at least one hardware processor, from each of the classified sentences of the email, at least one entity;
- and
- classifying, by the at least one hardware processor, based on the at least one extracted entity, the email.
A non-transitory computer readable medium having stored thereon machine readable instructions, the machine readable instructions, when executed by at least one hardware processor, cause the at least one hardware processor to:
- receive an email thread that includes a plurality of emails including an email that is to be classified, wherein the email thread pertains to a process that defines a logical path of contents of the email thread;
- identify for the email thread, the process associated with the email thread;
- select one or more classifiers of a plurality of classifiers based on transition probabilities of occurrences of the process that are affected by the reception of the email to be classified;
- determine, for each sentence of the email that includes a plurality of sentences, a corresponding classifier from the one or more classifiers that include at least two of a question, a statement, a command, or a request;
- apply, to each sentence of the email, the corresponding classifier to classify each sentence of the email to generate classified sentences;
- and
- classify, based on the classified sentences, the email.
Dependencies are inferred from the claim text.