Identifying changes in user characteristics using natural language processing
AssigneePaypal, Inc.
InventorsTomer Handelman, lulian-Corneliu-Ran Locar
Life of this patent
Estimating the term…
Abstract
Methods and systems for identifying characteristics of a user (e.g., a seller) based on Natural Language Processing (NLP). Transaction data of buyers may be collected to generate a sequence paragraph of seller name information for each buyer. NLP techniques such as word2vec may be used to vectorize the seller name information to determine relationships between sellers. Industry information may be determined using the vectors. Reliability checks may be performed to determine whether the data is robust to label the determined data.
Claims20
A system comprising:
- a non-transitory memory storing instructions;
- and
- one or more hardware processors coupled to the non-transitory memory and configured to read the instructions from the non-transitory memory to cause the system to perform operations comprising:
- receiving transaction log data associated with a plurality of users, the transaction log data comprising a plurality of sellers with whom the plurality of users have transacted;
- accessing a plurality of buyer sequences based on the transaction log data, each of the buyer sequences indicating a series of one or more purchases made by a respective user of the plurality of users with one or more of the plurality of sellers;
- mapping, based on a natural language processing algorithm, each of the plurality of sellers to a vector in a vector space;
- determining, for a first seller of the plurality of sellers, other sellers in the vector space within a threshold distance to the first seller;
- and
- determining a classification of the first seller based on classifications of the other sellers within the threshold distance of the first seller.
A method comprising:
- receiving, by one or more hardware processors, transaction log data associated with a plurality of users, the transaction log data comprising a plurality of sellers with whom the plurality of users have transacted;
- creating a plurality of buyer sequences based on the transaction log data for each of the plurality of users, each of the buyer sequences indicating a series of one or more purchases made by a respective user of the plurality of users at one or more of the plurality of sellers;
- mapping each of the plurality of sellers to a vector in a vector space using a natural language processing algorithm applied to the plurality of buyer sequences;
- determining, for a seller of the plurality of sellers, other sellers in the vector space within a threshold distance to the seller;
- and
- determining a classification of the seller based on classifications of the other sellers in the vector space within the threshold distance to the seller.
A non-transitory machine-readable medium having stored thereon machine-readable instructions executable to cause a machine to perform operations comprising:
- receiving transaction log data associated with a plurality of users, the transaction log data comprising a plurality of sellers with whom the plurality of users have transacted;
- accessing a plurality of buyer sequences based on the transaction log data, each of the buyer sequences indicating a series of one or more purchases made by a respective user of the plurality of users at one or more of the plurality of sellers;
- mapping, based on a natural language processing algorithm, each of the plurality of sellers to a vector in a vector space;
- determining for each of the plurality of sellers whether classification data is missing;
- determining, for each seller that classification data is determined to be missing, other sellers in the vector space within a threshold distance;
- and
- determining, for each seller of the plurality of sellers that classification data is determined to be missing, a classification based on classifications of ones of the plurality of sellers where classification data is not missing within the threshold distance of each seller.
Dependencies are inferred from the claim text.