A non-transitory computer-readable storage medium storing computer programs which, when executed by a processor of a computer equipment, cause the computer equipment to carry out actions, comprising:
- obtaining pending-transaction business data sent by a first user terminal (UE) corresponding to a first object, obtaining a predicted transaction price for the pending-transaction business data by performing price prediction on the pending-transaction business data, generating an initial transaction contract according to the predicted transaction price and the pending-transaction business data, and sending the initial transaction contract to the first UE, the first object being an invoice recipient or a payer of a transaction;
- obtaining feedback information from the first UE, predicting, according to the feedback information, contract update data that meets a transaction-success-rate condition, and obtaining an updated transaction contract by updating the initial transaction contract according to the contract update data, the feedback information comprising feedback of the first object on the predicted transaction price in the initial transaction contract, and the transaction-success-rate condition representing a condition for the probability of reaching a transaction with the first object;
- generating a transaction order containing the updated transaction contract and a contract signature generated for the updated transaction contract, and simultaneously sending the transaction order to the first UE and sending first invoicing critical data in the transaction order to an invoicing node in blockchain, the first invoicing critical data comprising all critical data for generating an electronic invoice, and being used for the invoicing node to generate a first block to-be-chained;
- in response to detecting that a second block to-be-chained has been created at the invoicing node, verifying whether a value of transaction funds in transaction data contained in the second block is identical with a value of contract funds in the updated transaction contract obtained by the computer equipment, and verifying whether a transaction signature contained in the second block is identical with the contract signature, wherein the second block to-be-chained is created in response to the transaction data and the transaction signature being received from the first UE, the transaction data indicates to transfer transaction funds in an account address of the first object to an account address of a main object corresponding to a computer equipment, and a parent block hash in a block header of the second block is a block hash of the first block;
- and
- in response to verifying that the value of the transaction funds in the transaction data is identical with the value of the contract funds in the updated transaction contract and the transaction signature is identical with the contract signature, sequentially chaining the first block and the second block, generating, according to the chained first block, a first electronic invoice corresponding to the transaction order and sending the first electronic invoice to the first UE, and updating, according to the chained second block, an account balance of the first object and an account balance of the main object in a state tree of the blockchain,
- wherein the computer programs executed by the processor to carry out actions of predicting, according to the feedback information, the contract update data that meets the transaction-completion-rate condition are executed by the processor to carry out actions, comprising:
- obtaining feedback semantic data by performing semantic analysis on the feedback information, and generating an object expected price according to the feedback semantic data;
- obtaining a price update parameter, and generating, according to the object expected price, the predicted transaction price, and the price update parameter, M transaction prices to-be-updated, wherein M is a positive integer;
- obtaining historical completed transaction behavior data corresponding to each of the M transaction prices to-be-updated, and obtaining M historical transaction behavior features by performing feature extraction on M historical completed transaction behavior data;
- obtaining a transaction-pending business feature by performing feature extraction on the pending-transaction business data, and obtaining M cross-attention scores by performing cross-attention processing on the transaction-pending business feature and the M historical transaction behavior features;
- determining an attention score threshold according to the transaction-success-rate condition, and forming a cross-attention sequence consisting of cross-attention scores greater than or equal to the attention score threshold among the M cross-attention scores, wherein the cross-attention sequence contains N cross-attention scores, and N is a positive integer less than or equal to M;
- and
- obtaining a transaction price to-be-updated corresponding to each of the N cross-attention scores, and determining the highest transaction price to-be-updated among N transaction prices to-be-updated as the contract update data.