Message recommendation using word isolation and clustering

Abstract

Network system provides a real-time adaptive recommendation set of documents with a high statistical measure of relevancy to the requestor device. The recommendation set is optimized based on analyzing text of documents of the interest set, categorizing these documents into clusters, extracting keywords representing the themes or concepts of documents in the clusters, and filtering a population of eligible documents accessible to the system utilizing site and or Internet-wide search engines. The system is either automatically or manually invoked and it develops and presents the recommendation set in real-time. The recommendation set may be presented as a greeting, notification, alert, HTML fragment, fax, voicemail, or automatic classification or routing of customer e-mail, personal e-mail, job postings, and offers for sale or exchange.

1 . A method for recommending email messages for further user action, the method comprising:
 storing a plurality of email messages in memory of a computing device, the email messages having been previously responded to by a user and the computing device including a processor and executable instructions stored in the memory; and   executing the instructions stored in the memory, wherein execution of the instructions by the processor:
 computes a plurality of similarity scores between the email messages previously responded to by the user, each of the similarity scores indicating a level of similarity between one or more words in a first email message from among the plurality of email messages and one or more words in a second email message from among the plurality of email messages, 
 groups the plurality of email messages into a plurality of clusters based on the computed similarity scores, and 
 recommends an email message received subsequent to the clustering of the email messages for further user action, wherein recommending the subsequently received email message includes calculating a relevance score for the subsequently received email message based on the plurality of clusters and one or more words in the subsequently received email message. 
   
     
     
         2 . The method of  claim 1 , wherein execution of the instructions by the processor further includes performing pre-processing on the email messages. 
     
     
         3 . The method of  claim 1 , wherein the preprocessing includes converting the email messages into a common format. 
     
     
         4 . The method of  claim 1 , wherein the preprocessing includes removing non keywords that do not facilitate grouping an email message into a particular group, and wherein the keywords that are not removed from the email message are used for computing the similarity score. 
     
     
         5 . The method of  claim 4 , wherein keywords for a particular group are based on frequency of a particular word appearing email messages of the group. 
     
     
         6 . The method of  claim 4 , wherein keywords for a particular group are based on a summary or concept of the group. 
     
     
         7 . The method of  claim 1 , wherein the groups of emails having a similar computed score correspond to an interest set from which a corresponding set of information, stored in a database, can be used for the recommendation. 
     
     
         8 . A system for recommending email messages for further user action, the system comprising:
 an assembly module that stores a plurality of email messages in memory of a computing device, the email messages having been previously responded to by a user and the computing device including a processor and executable instructions stored in the memory;   a pre-processing module that computes a plurality of similarity scores between the email messages previously responded to by the user, each of the similarity scores indicating a level of similarity between one or more words in a first email message from among the plurality of email messages and one or more words in a second email message from among the plurality of email messages,   a clustering module that groups the plurality of email messages into a plurality of clusters based on the computed similarity scores, and   a recommendation module that recommends an email message received subsequent to the clustering of the email messages for further user action, wherein recommending the subsequently received email message includes calculating a relevance score for the subsequently received email message based on the plurality of clusters and one or more words in the subsequently received email message.   
     
     
         9 . The system of  claim 8 , wherein the pre-processing module further performs one or more pre-processing processes on the email messages. 
     
     
         10 . The system of  claim 8 , wherein the pre-processing processes include converting the email messages into a common format. 
     
     
         11 . The system of  claim 8 , wherein the pre-processing processes include removing non keywords that do not facilitate grouping an email message into a particular group, and wherein the keywords that are not removed from the email message are used for computing the similarity score. 
     
     
         12 . The system of  claim 11 , wherein keywords for a particular group are based on frequency of a particular word appearing email messages of the group. 
     
     
         13 . The system of  claim 11 , wherein keywords for a particular group are based on a summary or concept of the group. 
     
     
         14 . The system of  claim 8 , wherein the groups of emails having a similar computed score correspond to an interest set from which a corresponding set of information, stored in a database, can be used for the recommendation. 
     
     
         15 . A non-transitory computer-readable storage medium, having embodied thereon a program executable by a processor to perform a method for recommending email messages for further user action, the method comprising:
 storing a plurality of email messages in memory of a computing device, the email messages having been previously responded to by a user and the computing device including a processor and executable instructions stored in the memory;   computing a plurality of similarity scores between the email messages previously responded to by the user, each of the similarity scores indicating a level of similarity between one or more words in a first email message from among the plurality of email messages and one or more words in a second email message from among the plurality of email messages;   grouping the plurality of email messages into a plurality of clusters based on the computed similarity scores; and   recommending an email message received subsequent to the clustering of the email messages for further user action, wherein recommending the subsequently received email message includes calculating a relevance score for the subsequently received email message based on the plurality of clusters and one or more words in the subsequently received email message.   
     
     
         16 . The non-transitory computer-readable storage medium of  claim 15 , wherein the method further includes performing pre-processing on the email messages. 
     
     
         17 . The non-transitory computer-readable storage medium of  claim 15 , wherein the preprocessing includes converting the email messages into a common format. 
     
     
         18 . The non-transitory computer-readable storage medium of  claim 15 , wherein the preprocessing includes removing non keywords that do not facilitate grouping an email message into a particular group, and wherein the keywords that are not removed from the email message are used for computing the similarity score. 
     
     
         19 . The non-transitory computer-readable storage medium of  claim 18 , wherein keywords for a particular group are based on frequency of a particular word appearing email messages of the group. 
     
     
         20 . The non-transitory computer-readable storage medium of  claim 18 , wherein keywords for a particular group are based on a summary or concept of the group. 
     
     
         21 . The non-transitory computer-readable storage medium of  claim 15 , wherein the groups of emails having a similar computed score correspond to an interest set from which a corresponding set of information, stored in a database, can be used for the recommendation.