Event-triggered patent dynamic valuation method and system

Abstract

The invention discloses an event-triggered patent dynamic valuation method and system, and the method comprises the steps: S1, monitoring an associated event of a target patent in real time; s2, when an event is monitored, resetting an evaluation weight matrix according to an event type; s3, dynamically adjusting the weight based on a time decay function; s4, adopting Monte Carlo simulation to calculate patent disposal depreciation rate distribution; and S5, generating a dynamic pledge rate suggestion. The technical problems of static estimation distortion, risk blind area disposal and high labor cost in the prior art are solved.

Description

Technical Field

[0001] The present invention relates to the intersection of financial technology and intellectual property evaluation, and in particular to an event-triggered patent dynamic valuation method and system, which is particularly suitable for real-time risk control in bank intellectual property pledge financing business. Background Art

[0002] Current problems with patent valuation include: static valuation distortion: traditional valuation methods use fixed weights and are unable to respond to technological iterations or legal emergencies; risk blind spots: the lack of tools to quantify the discount rate after patent defaults has led to a general bank pledge rate below 30% (equipment pledge rate reaches 70%); high labor costs: the cost of a single assessment is greater than 10,000 yuan, violating the requirement of the "New Science and Technology Finance Regulations" (2025) that "loans below 10 million yuan are exempt from third-party assessments." Summary of the Invention

[0003] In view of the shortcomings of the existing technology, the present invention provides an event-triggered patent dynamic valuation method and system to solve the technical problems of static valuation distortion, risk blind spots and high labor costs in the existing technology.

[0004] In order to solve the above technical problems, the present invention adopts the following technical solution: an event-triggered patent dynamic valuation method, characterized by comprising the following steps: S1. Real-time monitoring of events related to the target patent, including patent litigation, technology substitution, and mergers and acquisitions; S2. When an event is detected, the evaluation weight matrix is reset according to the event type, where: Patent litigation events increase the legal risk weight to 2.5 times the original value. The technology replacement event reduces the technology value weight to 0.7 times the original value. M&A events increase the market factor weight to 1.8 times its original value; S3. Dynamically adjust weights based on time decay function: Wt = e^(-λ•Δt) Where Wt represents the dynamic weight value after the event occurs, λ is the attenuation coefficient in the technology field (λ≥0.15 in the fields of artificial intelligence and blockchain, λ≤0.05 in the field of mechanical manufacturing), and Δt is the time difference between the current time and the event occurrence (unit: month); S4. Monte Carlo simulation is used to calculate the distribution of patent disposal discount rates: Generate market heat coefficient ξ ~ N(μ,σ²) and technology elimination probability ρ ~ Beta(α,β), Iterate the discount rate 10,000 times: Discount rate = basic discount rate × (1 + 0.3 × number of lawsuits) × (1 + ln(1+ρ)), Output the median discount rate and its 90% confidence interval. The 90% confidence interval is chosen based on historical data from the Beijing Intellectual Property Exchange, which shows that this interval covers 89.7% of actual auction results. S5. Generate dynamic pledge rate suggestions: Recommended pledge rate = [lower limit of patent valuation × (1 - median discount rate)] / loan amount × 100%.

[0005] Furthermore, the 2.5-fold adjustment coefficient for the patent litigation event is derived from the demand for compensation for value reduction corresponding to the average winning rate of 21.3% in patent infringement litigation; the 0.7-fold adjustment coefficient for the technology substitution event is set based on the 76% probability of patent depreciation within three years when the WIPO technology substitution index is greater than 0.8; the 1.8-fold adjustment coefficient for the M&A transaction event matches the value revaluation ratio of 178% of the average premium of patent M&A transactions.

[0006] Furthermore, the technology elimination probability ρ obeys Beta distribution, where the parameters in the field of artificial intelligence are α=2, β=5; and the parameters in the field of mechanical manufacturing are α=5, β=2.

[0007] Furthermore, the basic discount rate is determined according to the following rules: When the remaining life of the patent is ≥ 5 years, the basic discount rate is 0.2. When the remaining life of the patent is less than 5 years, the basic discount rate is 0.3. The remaining life of the patent is calculated as follows: the patent authorization date plus 20 years minus the current date.

[0008] The technical solution adopted by the present invention also includes an event-triggered patent dynamic valuation system, which is characterized by including: The event monitoring unit is configured to connect to the National Intellectual Property Administration API and the Judgment Document Network to capture patent litigation events in real time. Dynamic Assessment Unit, including: The weight reset module (executes step S2 of claim 1), connects to the weight rule library, stores a mapping table of event types and adjustment coefficients, the time decay engine (executes step S3 of claim 1), performs Wt = e^(-λ•Δt) weight calculation, the Monte Carlo simulator (executes step S4 of claim 1), generates a discount rate distribution based on historical auction data, and the output unit (executes step S5 of claim 1), generates a structured report containing valuation ranges and pledge rate recommendations.

[0009] Furthermore, the event monitoring unit further includes: The technology replaces the tracking subunit, which crawls the citation network data of the Derwent patent database; The transaction scanning sub-unit monitors merger and acquisition announcements from the Beijing, Shanghai and Guangzhou equity exchanges.

[0010] Furthermore, the output unit integrates a SHAP interpretable module to generate a risk factor contribution report.

[0011] Compared with the prior art, the present invention has the following beneficial effects: This invention discloses an event-triggered dynamic patent valuation method and system, eliminating the financial risks caused by patent valuation lags, quantifying the discount rate for patent breach settlements, and reducing valuation costs. Traditional models use linear superposition (discount rate = a × number of lawsuits + b), failing to couple legal and technical risks. This invention employs a multiplication structure: (1 + 0.3 number of lawsuits) to quantify legal losses, and (1 + ln(1 + ρ) to suppress technical extremes, reducing the error rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] The drawings are only for purposes of illustrating particular embodiments and are not to be considered limiting of the invention.

[0013] Figure 1 is a flow chart of a patent dynamic valuation method according to an embodiment of the present invention; Figure 2 is a flow chart of a patent dynamic valuation system according to an embodiment of the present invention; In the figure: 101-event monitoring unit, 1011-litigation capture subunit, 1012-technology alternative tracking subunit, 1013-transaction scanning subunit, 102-dynamic evaluation unit, 1021-weight reset module, 1022-time decay engine, 1023-Monte Carlo simulator, 103-output unit. DETAILED DESCRIPTION

[0014] The present invention is further described below with reference to the accompanying drawings. Implementation methods of the present invention include but are not limited to the following examples.

[0015] The following combination Figure 1, model operation process: event monitoring 101 (collecting data from the Judgment Document Network, Derwent Patent Library and Property Rights Exchange for event matching, and outputting event data packets); weight reset module 1021 receives the event data packet output by the event monitoring unit 101, queries the weight rule library according to the event type, obtains the adjustment coefficient (litigation x2.5, technology replacement x0.7), and performs weight reset (such as when a litigation event is detected, the legal risk weight = original value x2.5): the reset weight is input into the time decay engine 1022, and the current weight is calculated according to the formula Wt = e^(-λ•Δt); the decayed weight is output to the Monte Carlo simulator 1023; the pledge rate is output.

[0016] In an AI patent litigation case, a patent infringement lawsuit was detected on January 1, 2025, and the evaluation time was April 1, 2025 (3 months after the incident).

[0017] An event-triggered patent dynamic valuation method and system is adopted, and the implementation steps are as follows: Step 1: Event monitoring (step S1) and weight reset (step S2) 1. Initial weighting matrix (baseline): Legal risk weight: 15%, Technology value weight: 40%, Market factor weight: 45%; 2. Event-triggered reset: Patent litigation event detected → trigger weight reset rule: Legal risk weight = 15% * 2.5 = 37.5%; Updated weight matrix: Legal risk 37.5%, Technical value 40%, Market factor 45%, Legal risk weight = 15% × 2.5 = 37.5%, Technology value weight = 40% (remains unchanged), Market factor weight = 45% (remains unchanged; Normalization factor = 37.5% + 40% + 45% = 122.5%. After normalization: Legal risk weight = 37.5% / 122.5% ≈ 30.61%, Technology value weight = 40% / 122.5% ≈ 32.65%, Market factor weight = 45% / 122.5% ≈ 36.73%.

[0018] Step 2 Time Decay Calculation (Step S3) (Legal Risk Weights Only) Wt = e^(-λ•Δt), where λ (AI domain) = 0.15, Δt = 3 months (from the event to the present time); Calculate Wt = e^(-0.15×3) ≈0.6376, After legal risk weight attenuation = 30.61% × 0.6376 ≈ 19.52%, Technology value weight remains unchanged = 32.65%, Market factor weight remains unchanged = 36.73%, At this time, the total weight matrix = 19.52% + 32.65% + 36.73% = 88.9%, which needs to be normalized: Normalization factor = 88.9%, after normalization: Legal risk weight = 19.52% / 88.9% ≈ 21.96%, Technology value weight = 32.65% / 88.9% ≈ 36.72%, Market factor weight = 36.73% / 88.9% ≈ 41.32%, In this way, the legal risk weight increased from the initial 15% to 30.61% (normalized) due to the event trigger, and then decayed to 21.96% (normalized) after 3 months, while other weights were adjusted accordingly.

[0019] It should be clarified that time decay only adjusts the weight component triggered by the event (that is, the weight of the event that occurred, in this case, the patent litigation event).

[0020] Step 3: Monte Carlo simulation of discount rate (step S4) Discount rate formula: Discount rate = basic discount rate × (1 + 0.3 × number of lawsuits) × (1 + ln(1+ρ)); Parameters: base discount rate: 0.2 (because the remaining life is > 5 years), number of lawsuits: 3, ρ: 0.35 (the probability of technological obsolescence, obtained through sampling); Calculation (substitute into the discount rate formula): Part 1: 1 + 0.3 × 3 = 1.9, Part 2: 1 + ln(1+0.35) = 1 + ln(1.35) ≈ 1.3001, Discount rate = 0.2 × 1.9 × 1.3001 ≈ 0.4940 (i.e. 49.40%); The Monte Carlo simulation was performed 10,000 times, and each time a different ρ (obeying the Beta(2,5) distribution) was drawn to obtain 10,000 discount rates, and then the median and 90% confidence interval were taken.

[0021] In this example, we assume that p=0.35 for one sampling, but the actual simulation requires 10,000 times, and the final output is the median (the average of the 5000th and 5001th values after all discount rates are sorted) and 5% Quantile, 95% quantile, assuming that after 10,000 simulations, the median discount rate is 49.40%, with a 90% confidence interval of [41.2%, 57.6%].

[0022] Step 4: Generate a pledge rate recommendation (Step S5) New valuation range: After updating the weight matrix, the system recalculates the patent valuation range based on the weight matrix. Specifically, the scores of the three dimensions of the patent (legal risk, technological value, and market factors) are weighted and summed with the corresponding weights to obtain the patent's comprehensive score. The valuation range is then output based on the mapping relationship between the score and value (obtained through historical data training). Assume that the lower limit of the patent valuation is 7.2 million yuan and the loan amount is 10 million yuan. Pledge rate formula: Pledge ratio = [valuation lower limit × (1 - median discount rate)] / loan amount × 100%; Substitute: Pledge rate = [720 × (1 - 0.494)] / 1000 × 100% = 36.432%; Therefore, the recommended pledge rate is 36.4%.

[0023] Step 5: Actual verification: Auction results of an intellectual property exchange: Predicted value: The median discount rate is 49.4%, and the 90% confidence interval is [41.2%, 57.6%].

[0024] Actual value: The median discount rate is 51.2%, and the 90% confidence interval is [43.5%, 59.1%] The final auction discount rate was 52.10%.

[0025] With this technical solution, the actual discount rate of 52.10% falls within the prediction range of [41.2%, 57.6%].

[0026] like Figure 2As shown, the present invention provides an event-triggered patent dynamic valuation system, including: an event monitoring unit 101, including a litigation capture sub-unit 1011 configured to connect to the API of the State Intellectual Property Office and the Judgment Document Network to capture patent litigation events in real time; a dynamic evaluation unit 102, including: a weight reset module 1021, connected to the weight rule library, a time decay engine 1022, performing Wt = e^(-λ•Δt) weight calculation, a Monte Carlo simulator 1023, generating a discount rate distribution based on historical auction data; an output unit 103, generating a structured report containing valuation ranges and pledge rate recommendations.

[0027] Furthermore, the event monitoring unit 101 further includes: a technology replacement tracking subunit 1012 for crawling the citation network data of the Derwent patent database; and a transaction scanning subunit 1013 for monitoring the merger and acquisition announcements of the equity exchanges in Beijing, Shanghai and Guangzhou.

[0028] Furthermore, the output unit 103 integrates a SHAP interpretable module to generate a risk factor contribution report.

[0029] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by any technician familiar with this technical field within the technical scope disclosed by the present invention should be covered by the scope of protection of the present invention.

Claims

1. An event-triggered patent dynamic valuation method, characterized by The following steps are involved: S1. Real-time monitoring of events related to the target patent, including patent litigation, technology substitution, and mergers and acquisitions; S2. When an event is detected, the evaluation weight matrix is reset according to the event type, where: Patent litigation events increase the legal risk weight to 2.5 times the original value. The technology replacement event reduces the technology value weight to 0.7 times the original value. M&A events increase the market factor weight to 1.8 times its original value; S3. Dynamically adjust weights based on time decay function: Wt = e^(-λ•Δt) Where Wt represents the dynamic weight value after the event occurs, λ is the attenuation coefficient in the technology field (λ≥0.15 in the fields of artificial intelligence and blockchain, λ≤0.05 in the field of mechanical manufacturing), and Δt is the time difference between the current time and the event occurrence (unit: month); S4. Monte Carlo simulation is used to calculate the distribution of patent disposal discount rates: Generate market heat coefficient ξ ~ N(μ,σ²) and technology elimination probability ρ ~ Beta(α,β), Iterate the discount rate 10,000 times: Discount rate = basic discount rate × (1 + 0.3 × number of lawsuits) × (1 + ln(1+ρ)), Output the median discount rate and 90% confidence interval; S5. Generate dynamic pledge rate suggestions: Recommended pledge rate = [lower limit of patent valuation × (1 - median discount rate)] / loan amount × 100%.

2. The event-triggered patent dynamic valuation method according to claim 1, characterized in that : The 2.5-fold adjustment factor for the aforementioned patent litigation events stems from the demand for compensation for value reduction corresponding to the average winning rate of 21.3% in patent infringement litigation. The 0.7 times adjustment coefficient for the technology substitution event is set based on the 76% probability of patent devaluation within three years when the WIPO technology substitution index is greater than 0.

8. The 1.8x adjustment coefficient of the aforementioned M&A transaction matches the 178% average premium revaluation ratio of patent M&A transactions.

3. The event-triggered patent dynamic valuation method according to claim 1, characterized in that: The technology elimination probability ρ obeys the Beta distribution, where the parameters in the field of artificial intelligence are α=2, β=5, and the parameters in the field of mechanical manufacturing are α=5, β=2.

4. The event-triggered patent dynamic valuation method according to claim 1, characterized in that: The basic discount rate is determined according to the following rules: When the remaining life of the patent is ≥ 5 years, the basic discount rate is 0.

2. When the remaining life of the patent is less than 5 years, the basic discount rate is 0.

3. The remaining life of the patent is calculated as follows: the patent authorization date plus 20 years minus the current date.

5. An event-triggered patent dynamic valuation system, characterized by include: An event monitoring unit (101), including a litigation capture sub-unit (1011) configured to connect to the National Intellectual Property Administration API and the Judgment Document Network to capture patent litigation events in real time; A dynamic evaluation unit (102) includes a weight reset module (1021) connected to a weight rule library, a time decay engine (1022) for performing a weight calculation of Wt = e^(-λ•Δt), and a Monte Carlo simulator (1023) for generating a discount rate distribution based on historical auction data; The output unit (103) generates a structured report including the valuation range and pledge rate recommendations.

6. The event-triggered patent dynamic valuation system according to claim 5, characterized in that: The event monitoring unit (101) further includes: The technology replacement tracking sub-unit (1012) crawls the citation network data of the Derwent patent database; the transaction scanning sub-unit (1013) monitors the merger and acquisition announcements of the Beijing, Shanghai and Guangzhou equity exchanges.

7. The event-triggered patent dynamic valuation system according to claim 5, characterized in that: The output unit (103) integrates the SHAP interpretable module to generate a risk factor contribution report.

Images