Paypal, Inc.

Rushik Navinbhai Upadhyay

Dhamodharan Lakshmipathy

Aditya Kaulagi

Nandhini Ramesh

Meng Shi

Nikhil Rao

Quan Jin Ferdinand Tang

Hewen Wang

Zhe Chen

Van Hoang Nguyen

Suraj Belavalgidad

Francesco Gelli

Vijayalakshmi Dharmalingam

Chaitanya Maddipati

Venkatesan Thirunaukkarasu

Amit Nahata

Systems/techniques for facilitating context-enhanced category classification are provided. In various embodiments, a system can access a first textual description of a product or service. In various aspects, the system can identify, via execution of named entity recognition, one or more keywords in the first textual description. In various instances, the system can access, from a set of queryable databases, one or more second textual descriptions that respectively correspond to the one or more keywords. In various cases, the system can generate, via execution of word embedding, a first numerical representation of the first textual description and one or more second numerical representations of the one or more second textual descriptions. In various aspects, the system can identify, via execution of a machine learning classifier, a category label for the product or service, based on the first numerical representation and the one or more second numerical representations.

Context-enhanced category classification to generate product category labels used to conduct electronic transactions based on regulatory restrictions

Techniques are disclosed relating to implementing a machine learning approach to cross-language translation and search. In certain embodiments, a method may include receiving a plurality of characters of a first language that are unsegmented and grouping the plurality of character into multiple groups. The method also includes determining a set of word tokens based on one or more transliterations of the multiple groups and one or more translations of the multiple groups to a second language. Further, the method includes generating one or more word token solution sets by querying an index file using the one or more word tokens. The method also includes determining whether the index file references an entity name corresponding to the plurality of characters of the first language based on comparing the one or more token solution sets with the index file.

Machine learning approach to cross-language translation and search

A method may include determining that input text data includes a first keyword from a first set of keywords. The method also includes determining a similarity between the input text data and a first stored text string that has previously been identified as a false positive match for the first keyword, and based on the similarity, generating a first false positive score corresponding to the input text data. Further, the method includes determining a number of keywords, from a second set of keywords, that are included in the input text data, and based on the number of keywords, generating a second false positive score corresponding to the input text data. The method also includes calculating a final false positive score corresponding to the input text data based on the first false positive score and the second false positive score.

Identifying false positives between matched words

Techniques are disclosed relating to scoring partial matches between words. In certain embodiments, a method may include receiving a request to determine a similarity between an input text data and a stored text data. The method also includes determining, based on comparing one or more words included in the input text data with one or more words included in the stored text data, a set of word pairs and a set of unpaired words. Further, in response to determining that the set of unpaired words passes elimination criteria, the method includes calculating a base similarity score between the input text data and the stored text data based on the set of word pairs. The method also includes determining a scoring penalty based on the set of unpaired words and generating a final similarity score between the input text data and the stored text data by modifying the base similarity score based on the scoring penalty.

Algorithm for scoring partial matches between words

Techniques are disclosed relating to determining a predicted time interval for querying a database beginning at a starting point in time to retrieve a specified number of records. A computer system receives a request for records from a database that stores timestamped records, where the request is for a specified number of records beginning at a starting point in time. The computer system then determines a predicted time interval for querying the database beginning at the starting point in time to retrieve the specified number of records, where the predicted time interval may be determined by a machine learning module that is trained using historical database transaction data. Such techniques may advantageously improve the efficiency of individual queries in fetching a desired amount of data from a database for applications.

Time-based data retrieval prediction

A system performs operations that include determining match text data based on a comparison between input text data and an index of keywords and identifying a set of match word tokens from the match text data. The operations further include determining one or more solution sets of input word tokens based on comparing the set of match word tokens and the input text data. Additionally, the operations include calculating a respective solution set score for each of the one or more solution sets of input word tokens based on the calculated similarity values. A match between the input text data and the match text data is determined based on the first solution set score satisfying a score threshold.

Searching free-text data using indexed queries

Techniques are disclosed relating to determining phonetic similarity using machine learning. The techniques include accessing training data that includes a first set of words of a native language and a second set of words corresponding to verified transliterations of the first set of words from the native language to a target language. Further, they include generating a set of new transliterations of the first set of words from the native language to the target language and storing comparison information based on a comparison between words from the second set of words and word from the set of new transliterations of the first set of words. Finally, a similarity score is determined between a first word of the target language and a second word of the target language based on the comparison information.

Growing community of inventors

San Jose, CA, United States of America

Rushik Navinbhai Upadhyay