Latent Intent Clustering in High Latent Spaces

Abstract

A method of latent intent clustering is provided. The method comprises encoding identified features in a number of electronic user reports in a database. A binary matrix is created, wherein each row of the binary matric represents a different report and each column represents a different available feature. A 1 is placed in each cell of the matrix that matches a feature present in a user report. Cosine similarities are calculated for the user reports, and a similarity matrix is created, wherein each row and column of the binary matrix represents a different report, and wherein the cosine similarities of the reports are placed in corresponding cells of the matrix. The reports are clustered according to the cosine similarities. Features common to reports in each cluster are identified, and an intent of each report cluster is labeled according to the common features.

What is claimed is: 
     
         1 . A computer-implemented method of latent intent clustering, the method comprising:
 using a number of processors to perform the steps of:
 encoding identified features in a number of electronic user reports in a database; 
 creating a binary matrix, wherein each row of the binary matric represents a different user report and each column of the binary matrix represents a different available feature, and wherein a 1 is placed in each cell of the binary matrix that matches a feature present in a user report; 
 calculating cosine similarities of user reports represented in the binary matrix; 
 creating a similarity matrix, wherein each row and column of the binary matrix represents a different user report, and wherein the cosine similarities of the user reports are placed in corresponding cells of the similarity matrix; 
 clustering the user reports in the similarity matrix into report clusters according to the cosine similarities; 
 identifying features common to user reports in each report cluster; and 
 labeling an intent of each report cluster according to the common features. 
   
     
     
         2 . The method of  claim 1 , further comprising:
 receiving input from a user to create a new user report;   determining an intent of the new user report according to features of the new user report;   identifying a report cluster labeled with an intent that matches the intent of the new user report; and   providing the user with suggested features for the new user report according to features in the identified matching report cluster.   
     
     
         3 . The method of  claim 1 , wherein clustering the user reports comprises:
 calculating an average cosine similarity for each user report represented in the similarity matrix; and   clustering together user reports that have cosine similarities greater than their respective average cosine similarities.   
     
     
         4 . The method of  claim 1 , wherein encoding the features of each user report comprises one-hot encoding the fields in the user report. 
     
     
         5 . The method of  claim 1 , wherein encoding the features of each user report comprises:
 decomposing equations of a number of filters in the user report to produce decomposed filters; and   one-hot encoding the decomposed filters.   
     
     
         6 . The method of  claim 1 , wherein encoding the features of each user report comprises:
 decomposing equations of a number of custom fields in the user report to produce decomposed custom fields; and   one-hot encoding the decomposed custom fields.   
     
     
         7 . The method of  claim 1 , wherein encoding the features of each user report comprises one-hot encoding a title and description. 
     
     
         8 . A system for latent clustering, the system comprising:
 a storage device configured to store program instructions; and   a number of processors operably connected to the storage device and configured to execute the program instructions to cause the system to:
 encode identified features in a number of electronic user reports in a database; 
 create a binary matrix, wherein each row of the binary matric represents a different user report and each column of the binary matrix represents a different available feature, and wherein a 1 is placed in each cell of the binary matrix that matches a feature present in a user report; 
 calculate cosine similarities of user reports represented in the binary matrix; 
 create a similarity matrix, wherein each row and column of the binary matrix represents a different user report, and wherein the cosine similarities of the user reports are placed in corresponding cells of the similarity matrix; 
 cluster the user reports in the similarity matrix into report clusters according to the cosine similarities; 
 identify features common to user reports in each report cluster; and 
 label an intent of each report cluster according to the common features. 
   
     
     
         9 . The system of  claim 8 , wherein the processors further execute instructions to:
 receive input from a user to create a new user report;   determine an intent of the new user report according to features of the new user report;   identify a report cluster labeled with an intent that matches the intent of the new user report; and   provide the user with suggested features for the new user report according to features in the identified matching report cluster.   
     
     
         10 . The system of  claim 8 , wherein clustering the user reports comprises:
 calculating an average cosine similarity for each user report represented in the similarity matrix; and   clustering together user reports that have cosine similarities greater than their respective average cosine similarities.   
     
     
         11 . The system of  claim 8 , wherein encoding the features of each user report comprises one-hot encoding the fields in the user report. 
     
     
         12 . The system of  claim 8 , wherein encoding the features of each user report comprises:
 decomposing equations of a number of filters in the user report to produce decomposed filters; and   one-hot encoding the decomposed filters.   
     
     
         13 . The system of  claim 8 , wherein encoding the features of each user report comprises:
 decomposing equations of a number of custom fields in the user report to produce decomposed custom fields; and   one-hot encoding the decomposed custom fields.   
     
     
         14 . The system of  claim 8 , wherein encoding the features of each user report comprises one-hot encoding a title and description. 
     
     
         15 . A computer program product for latent intent clustering, the computer program product comprising:
 a computer-readable storage medium having program instructions embodied thereon to perform the steps of:
 encoding identified features in a number of electronic user reports in a database; 
 creating a binary matrix, wherein each row of the binary matric represents a different user report and each column of the binary matrix represents a different available feature, and wherein a 1 is placed in each cell of the binary matrix that matches a feature present in a user report; 
 calculating cosine similarities of user reports represented in the binary matrix; 
 creating a similarity matrix, wherein each row and column of the binary matrix represents a different user report, and wherein the cosine similarities of the user reports are placed in corresponding cells of the similarity matrix; 
 clustering the user reports in the similarity matrix into report clusters according to the cosine similarities; 
 identifying features common to user reports in each report cluster; and 
 labeling an intent of each report cluster according to the common features. 
   
     
     
         16 . The computer program product of  claim 15 , further comprising instructions for:
 receiving input from a user to create a new user report;   determining an intent of the new user report according to features of the new user report;   identifying a report cluster labeled with an intent that matches the intent of the new user report; and   providing the user with suggested features for the new user report according to features in the identified matching report cluster.   
     
     
         17 . The computer program product of  claim 15 , wherein clustering the user reports comprises:
 calculating an average cosine similarity for each user report represented in the similarity matrix; and   clustering together user reports that have cosine similarities greater than their respective average cosine similarities.   
     
     
         18 . The computer program product of  claim 15 , wherein encoding the features of each user report comprises one-hot encoding the fields in the user report. 
     
     
         19 . The computer program product of  claim 15 , wherein encoding the features of each user report comprises:
 decomposing equations of a number of filters in the user report to produce decomposed filters; and   one-hot encoding the decomposed filters.   
     
     
         20 . The computer program product of  claim 15 , wherein encoding the features of each user report comprises:
 decomposing equations of a number of custom fields in the user report to produce decomposed custom fields; and   one-hot encoding the decomposed custom fields.   
     
     
         21 . The computer program product of  claim 15 , wherein encoding the features of each user report comprises one-hot encoding a title and description.