91 results about "Fraudulent transaction" patented technology

Embodiments provide artificial intelligence methods and systems for evaluating vulnerability risks of issuer authorization system. Method performed by a server system includes accessing a set of payment transaction data including subset of fraudulent transaction data. Method includes generating via a machine learning model, set of synthetic transaction data based on the subset of fraudulent transaction data. Method includes accessing set of historical card velocity features and collating the set of synthetic transaction data and the set of historical card velocity features to generate set of enriched synthetic transaction data. Method includes extracting via a classifier, subset of feasible fraudulent transaction data from the set of enriched synthetic transaction data. Method includes generating simulated authorization model based on the set of payment transaction data. Method includes classifying via simulated authorization model, each enriched synthetic transaction from the subset of feasible fraudulent transaction data as one of fraudulent transaction and non-fraudulent transaction.

In particular embodiments, techniques include a computer-implemented method including monitoring a plurality of real-time streams of content and a plurality of transactions in real-time. A machine-learning model trained to detect fraudulent transaction activity is used to determine that one or more of the real-time streams are associated with fraudulent transaction activity based on input corresponding to the monitored real-time streams and plurality of transactions. A request is received from a user device of a first user to initiate a transaction on behalf of the first user and it is determined that a characteristic of the requested transaction is indicative of a fraudulent transaction associated with the fraudulent transaction activity associated with one or more of the real-time streams. In response, processing of the requested transaction is modified.

A system includes one or more processors to generate a node graph; determine a first node of the node graph comprises a fraudulent flag indicating a first entity of the first node facilitated a fraudulent transaction; responsive to the determination, identify a set of nodes of the node graph responsive to each node of the set of nodes having a direct transaction edge connection with the first node within the node graph or an indirect transaction edge connection with the first node via at least one node directly connected with the first node; generate a fraudulent flag in a subset of the set of nodes responsive to determining each node in the subset satisfies a matching policy; and generate a record identifying the first node and each node of the subset of nodes.

The technology described herein relates to systems, methods, and computer storage media, among other things, for determining whether an electronic transmission (e.g., associated with an electronic payment transaction) should be blocked (e.g., based on being a fraudulent transaction). In embodiments, a policy-based reinforcement learning risk decision agent is used to make these determinations for a plurality of stages associated with the electronic payment transaction (e.g., a pre-authorization stage, a post-authorization stage, and a delay-captured stage). The policy-based reinforcement learning risk decision agent can be trained using previous electronic payment transaction data for previous electronic payment transactions. For example, this particular agent can be trained using pre-authorization electronic payment transaction data, post-authorization electronic payment transaction data, and delay-captured electronic payment transaction data for each of the previous electronic payment transactions.

There is provided a computing system for optimizing a plurality of fraud detection strategies used to generate a corresponding set of potentially fraudulent transactions. The system determines an overall fraud value such as an average fraud value for each transaction based on pre-defined factors and identifies a particular strategy having a highest average fraud value for its fraudulent transactions as a highest priority on a ranked list of strategies. The system is configured to remove each transaction from the remaining other strategies if the same as the fraudulent transactions in the identified strategy and calculate an average fraud value for the remaining other strategies. The system then ranks the next highest priority fraud detection strategy having the highest average fraud value while removing its corresponding transactions flagged from other remaining strategies and repeat the ranking until all the strategies have been ranked and apply the ranked list to subsequent transactions.

The invention relates to the technical field of financial risk management, and discloses a risk control method and system for identifying a transition financial business greening risk, and the method comprises the steps: deconstructing an enterprise commitment into an expected economic behavior vector through a transition knowledge graph, capturing an actual economic behavior data flow, and carrying out the logic consistency matching; according to the method, environmental benefit verification is converted into economic behavior logic chain review, so that a financial institution can avoid high-cost environmental auditing to return to business essence judgment, and meanwhile, bogus transactions are effectively identified through business activeness analysis of a transaction counterparty; and the risk identification capability of the transformed financial business is obviously improved.

The invention relates to the field of information security. A method for detecting fraudulent user transactions comprises the steps of: obtaining a sequence of transactions made by a user within a given time frame, each transaction being characterized by a set of attributes; converting each obtained transaction into an attribute vector on the basis of the corresponding set of transaction attributes; processing the obtained attribute vectors with the aid of a coding machine learning model, with a representation of a sequence of attribute vectors being produced in the form of a hidden state matrix; processing the hidden state matrix with the aid of a generative machine learning model (MLM) trained on hidden state matrices of sequences of transaction attribute vectors, with the generative MLM generating from the hidden state matrix interrelated attribute vectors of one or several transactions; generating the last transaction of the user on the basis of the generated attribute vector; comparing the last transaction made by the user and the generated last transaction of the user by calculating an anomaly assessment of the transaction made; signalling a fraudulent transaction if the anomaly assessment value is above a threshold value. The invention provides for more complete and accurate detection of fraudulent transactions.

A system is adapted to automatically identify suspected fraudulent transactions. The system includes a fraud management server configured to perform these operations: receiving unlabeled transactions, each having a number of features, and storing them in a transaction repository; with the features, determining a risk score for each transaction; based on the risk scores, dividing the unlabeled transactions into bins in order of their risk scores; labeling transactions of the first bin legitimate and those of last bin as fraudulent; with the labeled transactions, training a first machine learning model; with the trained first machine learning model, labeling transactions of a second bin and a second-to-last bin as either fraudulent or legitimate; storing the labeled transactions of the first bin, second bin, second-to-last bin, and last-bin in the transaction repository; and with the labeled transactions of the first bin, second bin, second-to-last bin, and last-bin, training a second machine learning model.

The invention discloses a commercial bank blacklist management method based on a big data platform and Elasticsearch, and the method comprises the steps: building a blacklist data center based on the big data platform, integrating multi-source heterogeneous data, and carrying out the real-time data cleaning and standardization processing; establishing a distributed real-time retrieval engine by utilizing Elasticsearch, and constructing a reverse index for the processed data; establishing an associated risk map model, dynamically analyzing the association degree between the client and the blacklist main body through a map calculation engine, generating a risk conduction coefficient, and storing the risk conduction coefficient; during business handling, a multi-condition combination query request is initiated to Elasticsearch through an ESB (Enterprise Service Bus) real-time calling interface; a dynamic updating mechanism is adopted, and the blacklist state is automatically updated when a risk threshold value is triggered. According to the method, by integrating multi-source heterogeneous data, millisecond risk interception is realized, a dynamic association graph is constructed, the active defense capability of a bank on risks such as fraudulent transactions, credit default and money laundering behaviors is improved, and the method is suitable for risk management and control of core business scenes such as pre-loan auditing, transaction monitoring and anti-money laundering.

The described technology relates to a deep learning based method and system for detecting abnormal cryptocurrency transaction between computing devices in a blockchain network. In once aspect, the method includes generating, at a server, a first data set, a second data set, and a third data set from the transactions of a user wallet address with at least one other user wallet address. The method may also include running a pre-learned deep learning module to extract a first feature vector, a second feature vector, and a third feature vector from the first data set, the second data set, and the third data set. The method may further include converting the first feature vector, the second feature vector, and the third feature vector into an intermediate value and comparing the intermediate value to a predetermined reference value to determine if a fraudulent transaction associated with the user wallet address has occurred.

A verified real estate registration system, where individuals may register, for avoiding a fraudulent transaction occurring with respect to real property. During registration, basic information identifying the individual or entity registering along with verifying documentation are entered into the system. Once the identity of the individual or entity is confirmed, the individual or entity then becomes a registered user of the system and is permitted to enter any real property, solely or jointly owned by the registered user, into a data base of the system. The system also permits registered users to generate a warning notice, with respect to real property, which is recordable in a chain of title of any real property, owned by the registered individual or entity, so as to avoid any fraudulent transfer occurring respect to the real property having the warning notice recorded in the chain of title.

Systems and processes are disclosed for real-time fraud detection and prevention in in-person transactions. The invention utilizes an AI / ML engine to analyze customer application data for inconsistencies and unusual requests indicative of potential fraud. Concurrently, a real-time conversation analysis engine with speech recognition algorithms monitors interactions between bank associates and customers, identifying suspicious speech patterns, hesitations, and keywords associated with scams. By combining insights from application data and conversational analysis, the system generates a comprehensive risk assessment. When a high probability of fraud is detected, an alert notifies the bank associate, security personnel, and other relevant individuals. This proactive approach enables immediate action to prevent fraudulent transactions, reducing manipulation risks and minimizing financial losses. The system continuously learns from new data, adapting to evolving fraud tactics, thus providing robust, long-term protection for financial institutions and their customers.

A method for detecting fraudulent transactions may include receiving transaction data for a plurality of transactions, assigning a unique identifier to each transaction, sorting the plurality of transactions into a first plurality of groups of transactions by a first sorting parameter, assigning the unique identifier of a first transaction within each group to all other transactions within the respective group, sorting the plurality of transactions into a second plurality of groups of transactions by a second sorting parameter, assigning the unique identifier of a second transaction within each group to all other transactions within the respective group, determining whether each group of transactions is a sequence of fraudulent transactions, and upon determining that a group of transactions is a sequence of fraudulent transactions, marking each transaction among the determined group of transactions as potentially fraudulent.

A system and method including: obtaining pending transactions; processing weightings for parameters associated with the pending transactions; encoding the pending transactions with the processed weightings into corresponding transaction vector representations; obtaining nearest transactions using the transaction vector representations; determining whether the pending transactions are flagged for association with at least one risky entity; in response to determining that the pending transactions are flagged for association with a risky entity, generating an alert for the nearest transactions; in response to determining that the pending transactions are not flagged, determining whether the nearest transactions are associated with at least one risky entity; in response to determining that the pending transactions or the nearest transactions are flagged for association with at least one risky entity, generating an alert for the pending transactions; and output the generated alerts to one or more computing devices via a network interface.

The present invention discloses a financial system monitoring system based on blockchain and cloud computing, belonging to the field of financial system monitoring technology. The system includes a data acquisition module, a data analysis module, a data processing module, a monitoring module, a prediction module, and a blockchain storage module; the output of the data acquisition module is connected to the input of the data analysis module, the monitoring module, and the blockchain storage module; the output of the data analysis module is connected to the input of the data processing module; the output of the data processing module is connected to the input of the monitoring module and the prediction module; and the output of the monitoring module is connected to the input of the blockchain storage module. The system can effectively prevent the occurrence of fraudulent transactions, coerced transactions, and other undesirable behaviors during financial transactions, and all data is uploaded to the blockchain storage module, further enhancing traceability and effectively preventing data loss.

The invention discloses a fraud risk prediction method and device in combination with historical behavior analysis, and relates to the technical field of transaction risk prediction.The method comprises the steps that when a user initiates a transaction, transaction state information is monitored, and the transaction state information comprises transaction behaviors, network behaviors and biological behaviors; counting accounts whose transaction frequency with the user is greater than or equal to a threshold value, and obtaining a user associated account set; if there is no risk account in the associated account set (i.e., the historical fraud frequency is lower than a threshold value), counting a fraud transaction proportion in historical transaction behaviors having the same modality as the transaction behavior and the network behavior; if the fraudulent transaction proportion is lower than a threshold value, counting historical biological behaviors of the user; and if the historical biological behavior is inconsistent with the current biological behavior, generating a fraud risk prompt and feeding back the fraud risk prompt to the user side, thereby achieving the technical effect of improving the accuracy and reliability of fraud risk prediction.

The invention discloses an unbalanced fraud detection method based on WGAN-GP oversampling, and relates to the technical field of information security detection, and the method comprises the following steps: judging an unbalanced transaction data set, if the number of fraud transactions is equal to the number of normal transactions, outputting the unbalanced transaction data set as a balanced transaction data set, and if the number of fraud transactions is equal to the number of normal transactions, outputting the balanced transaction data set; otherwise, calculating the information value of the unbalanced transaction data set, training the WGAN-GP based on the information value, and generating a fraudulent transaction sample; training a strong classifier through an ensemble learning method, calculating a judgment condition for judging that a sample is a fraudulent transaction, carrying out fraudulent transaction integrated screening on the fraudulent transaction, supplementing qualified generated fraudulent transactions into an unbalanced transaction data set, repeating the operation until the number of the fraudulent transactions is equal to the number of normal transactions, and outputting a balanced transaction data set; noise is effectively suppressed while high-quality fraud samples are generated, and the precision and generalization ability of unbalanced fraud detection are remarkably improved.

The invention provides a payment fraud real-time identification and disposal method and system based on machine learning, and relates to the technical field of computers, and the method comprises the steps: constructing a transaction relation graph, recognizing an abnormal transaction mode through a graph attention network, and calculating a transaction suspicious degree score; and when the preset threshold value is exceeded, triggering a self-adaptive verification strategy based on the knowledge graph and reinforcement learning, executing a corresponding verification means and processing a transaction result. According to the invention, accurate identification and efficient disposal of fraudulent transactions can be realized, the payment security is improved, the error interception rate is reduced, and the user experience is optimized.

Systems and methods for evaluating risks for physical locations are disclosed. A request for risk evaluation of transactions associated with a physical location located within a predetermined region including a plurality of neighboring locations is received. A spatial model is implemented to generate a first risk score based on chargeback data associated with the physical location and chargeback data of the plurality of neighboring locations. A temporal model is implemented to generate a second risk score for the physical location. A final risk score for the physical location is generated based on the first risk score and the second risk score. The final risk score is transmitted to a computing device for detecting fraudulent transactions associated with the physical location in response to the request.

The invention discloses a fraudulent transaction detection method and device, equipment and a medium, which are used for improving the understanding capability of a transaction detection model on transaction data, so that the fraudulent transaction detection effect is improved, and the situations of missing detection, false alarm and the like are avoided. The fraudulent transaction detection method provided by the invention is suitable for the field of bank card payment, and the method comprises the steps: obtaining original transaction data; wherein the original transaction data is presented in the form of coded fields, and comprises a plurality of coded fields of at least one type; based on a preset domain knowledge base, converting the original transaction data into a transaction text described by adopting a natural language; wherein the domain knowledge base comprises a corresponding relation between a coding field and a semantic text described by adopting a natural language; and based on the transaction text described by adopting the natural language, adopting at least one preset transaction detection model to carry out fraudulent transaction detection.

Methods, systems, and apparatuses are described herein for improving computer authentication processes through excluding fraudulent transactions in transaction-based authentication. A computing device may receive a request for access to an account from a user. The computing device may provide transaction data to a machine learning model. The computing device may receive data indicating a confidence threshold associated with the user from the machine learning model. The computing device may generate a modified set of merchant choices for the user by excluding merchants related to potentially fraudulent transactions within a predetermined time period. An authentication question may be generated, and access to the account may be provided based on a response to the authentication question.

Methods, systems, and apparatuses are described herein for improving computer authentication processes through excluding fraudulent transactions in transaction-based authentication. A computing device may receive a request for access to an account from a user. The computing device may provide transaction data to a machine learning model. The computing device may receive data indicating a confidence threshold associated with the user from the machine learning model. The computing device may generate a modified set of merchant choices for the user by excluding merchants related to potentially fraudulent transactions within a predetermined time period. An authentication question may be generated, and access to the account may be provided based on a response to the authentication question.

The invention provides a promotion income sharing method, and relates to the field of intelligent wardrobes, and the method comprises the steps: responding to an exchange request of a digital exchange coupon of a user side, and generating an encryption certificate containing an effective timestamp and a unique identifier based on a first node; when a corresponding target device identifies the digital exchange coupon, the encrypted certificate is verified through a second node; if the encrypted certificate is in an unused state and is valid, controlling the target device to execute an exchange request of the digital exchange certificate, determining a promotion income amount corresponding to the exchange request, and writing corresponding verification success information and the promotion income amount into the block chain; and if the encryption certificate is in the expired state and / or the used state, writing verification failure information of the encryption certificate into the block chain. By means of block chain distributed verification and non-tampering characteristics, the states of the exchange coupons are clearly distinguished, the problem of state fuzziness is solved, it is ensured that the basis for promotion income calculation is real and reliable, and fraudulent transactions are reduced.

A fraud detection platform may receive transaction data relating to a transaction conducted by a user with a transaction card. The fraud detection platform may receive, from a biometric sensor of the transaction card, biometric data relating to one or more biometric characteristics of the user during the transaction. The fraud detection platform may receive, from an accelerometer of the transaction card, card movement data relating to a measure of shaking of the transaction card by the user during the transaction. The fraud detection platform may process the transaction data, the biometric data, and the card movement data, with a fraud detection model, to determine a fraud score associated with the transaction. The fraud detection platform may perform one or more actions based on the fraud score.

Apparatus and methods for an opt-in distributed ledger consortium to detect and prevent fraudulent transactions are provided. Two or more entities may opt into a private distributed ledger. A program may receive financial transaction information. The program may record the transaction on the distributed ledger. The program may activate a smart application on the ledger. The smart application may analyze each transaction for indicators of illegal activity. When indicators of illegal activity are found within a transaction, the program may generate a report. The program may transmit the report to each entity that has opted into the private distributed ledger. The program may also record the report on the private distributed ledger.

A computing platform is configured to (i) identify a candidate set of card-not-present (CNP) transactions that are candidates for potential involvement in fraudulent activity, (ii) for each respective CNP transaction in the candidate set, determine a respective combination of transaction-element values for a set of transaction elements comprising at least (a) a first transaction element indicating a Bank Identification Number (BIN) number of a respective PAN involved in the respective CNP transaction, (b) a second transaction element indicating a client involved in the respective CNP transaction, and (c) a third transaction element indicating a merchant involved in the respective CNP transaction, (iii) based on an evaluation of the respective combinations of transaction-element values determined for the CNP transactions in the candidate set, identify at least one at-risk combination of transaction-element values that is associated with a risk of fraudulent activity, and (iv) use the identified at least one at-risk combination of transaction-element values as a basis for deploying logic for identifying CNP transactions that present a risk of fraudulent activity.

A computing platform is configured to (i) identify a candidate set of card-not-present (CNP) transactions that are candidates for potential involvement in fraudulent activity, (ii) for each respective CNP transaction in the candidate set, determine a respective combination of transaction-element values for a set of transaction elements comprising at least (a) a first transaction element indicating a Bank Identification Number (BIN) number of a respective PAN involved in the respective CNP transaction, (b) a second transaction element indicating a client involved in the respective CNP transaction, and (c) a third transaction element indicating a merchant involved in the respective CNP transaction, (iii) based on an evaluation of the respective combinations of transaction-element values determined for the CNP transactions in the candidate set, identify at least one at-risk combination of transaction-element values that is associated with a risk of fraudulent activity, and (iv) use the identified at least one at-risk combination of transaction-element values as a basis for deploying logic for identifying CNP transactions that present a risk of fraudulent activity.

This invention provides a machine learning-based method and system for real-time identification and handling of payment fraud, relating to the field of computer technology. The method includes constructing a transaction relationship graph, using graph attention networks to identify abnormal transaction patterns, and calculating transaction suspiciousness scores. When a preset threshold is exceeded, an adaptive verification strategy based on knowledge graphs and reinforcement learning is triggered to execute corresponding verification measures and process the transaction results. This invention enables accurate identification and efficient handling of fraudulent transactions, improves payment security, reduces false interception rates, and optimizes user experience.

In point-to-point payment, fraudulent transactions are detected without an intermediary. An audit trace associated with the transaction chain is generated for transferring the digital currency from the initial sending wallet to the final receiving wallet and associating the audit trace with the token of the digital currency. Before being received by the final receiving wallet, each token maintains an audit trace of an intermediate digital wallet in which the token is transacted. A request is received to identify whether a first transaction in a transaction chain is a fraudulent transaction. In response to the request, a first audit trace associated with a first token associated with the first transaction is retrieved. The first transaction is determined as a fraudulent transaction when a wallet clone or value tampering is detected in a first token associated with the first transaction by analyzing the first audit trace.

The present invention relates to a transaction anti-fraud method based on big data, and relates to the field of security technology. The present application uses transaction big data to create a directed transaction graph that describes transaction behaviors between accounts. The nodes in the directed transaction graph represent the accounts involved in various transactions, the direction of the directed edges represents the flow direction of the transaction amount, and the weight corresponding to the directed edges represents the transaction amount between the account nodes. Within a set time slice, a transaction network is constructed between any two accounts to be analyzed, formed by transactions through intermediate accounts. For any transaction network, the transaction network features of each transaction network in different time slices are arranged in sequence to form a time-series transaction network feature. A fraud recognition model is constructed and trained using the labeled transaction big data, aiming to automatically identify the difference between normal transactions and suspected fraudulent transactions through the fraud recognition model to identify suspicious transaction behaviors in transactions. The fraud recognition model includes: a graph neural network, a recurrent neural network, and a classification head.