Systems and methods for performing a computer-implemented prior art search

Abstract

In some embodiments, a computer-implemented method for retrieving a similar document from a corpus of documents is provided. The method may include receiving a search document comprising a set of words; applying a first encoder to generate a first vector; applying a second encoder to generate a second vector; determining a first similarity between the first vector of the search document and the first vector of each document of the corpus of documents; determining a second similarity between the second vector of the search document and the second vector of each document of the corpus of documents; generating a first ranked list of documents based on the first similarity; generating a second ranked list of documents based on the second similarity; applying a voting algorithm to determine a score associated with each document; and outputting a third ranked list of documents based on the determined score.

1 . A computer implemented method for generating a document database, the method comprising:
 receiving a document of a plurality of documents, the document comprising a set of words;   applying a first encoder to the set of words to generate a first vector;   applying a second encoder to the set of words to generate a second;   indexing the document using the first vector and the second vector into a searchable index; and   enabling searching for the document using the index.   
     
     
         2 . The method of  claim 1 , wherein the index comprises a first index for the first vector and second index for the second vector. 
     
     
         3 . The method of  claim 1 , further comprising:
 applying a model to the set of words to generate one or more tags associated with the document; and   based on the one or more tags, updating a node-edge graph with the document, wherein the node-edge graph represents relationships between the plurality of documents.   
     
     
         4 . The method of  claim 1 , further comprising:
 parsing the document to identify a trigger word, the trigger word indicating a subset of words;   determining that the document does not contain the trigger word; and   storing the document in an exception database.   
     
     
         5 . The method of  claim 1 , wherein the first encoder comprises a neural network encoder. 
     
     
         6 . The method of  claim 1 , wherein applying the second encoder comprises parsing the set of words and removing occurrences of a stop word from the set of words to generate a subset of words. 
     
     
         7 . The method of  claim 6 , further comprising:
 determining the number of times each of the subset of words appears in the subset of words; and   determining a rarity of each of the subset of words.   
     
     
         8 . The method of  claim 7 , wherein the rarity comprises a number of times a word appears in the document compared to a number of times the word appears in the plurality of documents. 
     
     
         9 . The method of  claim 3 , wherein the one or more tags belong to a set of tags, the set of tags being generated by a model applied to the plurality of documents. 
     
     
         10 - 20 . (canceled) 
     
     
         21 . A system for generating a document database, the system comprising:
 a memory storing instructions; and   one or more processors configured to execute the instructions to perform operations comprising:
 receiving a document of a plurality of documents, the document comprising a set of words; 
 applying a first encoder to the set of words to generate a first vector; 
 applying a second encoder to the set of words to generate a second; 
 indexing the document using the first vector and the second vector into a searchable index; and 
 enabling searching for the document using the index. 
   
     
     
         22 . The system of  claim 21 , wherein the index comprises a first index for the first vector and second index for the second vector. 
     
     
         23 . The system of  claim 21 , the operations further comprising:
 applying a model to the set of words to generate one or more tags associated with the document; and   based on the one or more tags, updating a node-edge graph with the document, wherein the node-edge graph represents relationships between the plurality of documents.   
     
     
         24 . The system of  claim 21 , the operations further comprising:
 parsing the document to identify a trigger word, the trigger word indicating a subset of words;   determining that the document does not contain the trigger word; and   storing the document in an exception database.   
     
     
         25 . The system of  claim 21 , wherein the first encoder comprises a neural network encoder. 
     
     
         26 . The system of  claim 21 , wherein applying the second encoder comprises parsing the set of words and removing occurrences of a stop word from the set of words to generate a subset of words. 
     
     
         27 . The system of  claim 26 , the operations further comprising:
 determining the number of times each of the subset of words appears in the subset of words; and   determining a rarity of each of the subset of words.   
     
     
         28 . The system of  claim 27 , wherein the rarity comprises a number of times a word appears in the document compared to a number of times the word appears in the plurality of documents. 
     
     
         29 . The system of  claim 23 , wherein the one or more tags belong to a set of tags, the set of tags being generated by a model applied to the plurality of documents.