KR-20260062868-A - APPARATUS, METHOD AND PROGRAM FOR CALCULATING PERFORMANCE REWARD BASED ON ARTIFICIAL INTELLIGENCE

Abstract

The AI-based performance reward calculation device includes: a data input unit that receives data of a predetermined Key Performance Indicator (KPI) for a plurality of evaluation subjects; a reward score calculation unit that calculates a predetermined performance-based score, deduction score, risk score, and sustainability score for each of the plurality of evaluation subjects by applying a weight and a predetermined normalization function according to the contribution of each indicator learned from past performance data through an AI learning model to the KPI data; and a differential reward amount calculation unit that calculates a differential reward amount for each of the plurality of evaluation subjects based on the reward score calculated by applying predetermined global weights learned through an AI learning model to each of the scores calculated for each of the plurality of evaluation subjects.

Inventors

• 이병갑
• 모선영
• 김도수

Assignees

• 애디플 주식회사

Dates

Publication Date

20260507

Application Date

20251027

Priority Date

20241025

Claims (16)

1. In an AI-based performance reward calculation device, A data input unit that receives specific Key Performance Indicator (KPI) data for multiple evaluation subjects; A reward score calculation unit that calculates a predetermined performance-based score, deduction score, risk score, and sustainability score for each of the plurality of evaluation subjects by applying a weight based on the contribution of each indicator learned from past performance data through an AI learning model to the above KPI data and a predetermined normalization function; and A differential compensation amount calculation unit that calculates a differential compensation amount for each of the plurality of evaluation subjects based on a compensation score calculated by applying predetermined global weights learned through an AI learning model to each of the scores calculated for each of the plurality of evaluation subjects; AI-based performance reward calculation device including
2. In Article 1, The KPI items of the above KPI data are Click-through rate (CTR); representing the ratio of the number of clicks to the total number of ad impressions. Conversion Rate (CVR); representing the percentage of members who click on an ad and subsequently achieve a specific target behavior. Return on Advertising Spend (ROAS), representing the ratio of total sales generated to advertising costs; and Cost Per Acquisition (CPA), which represents the average cost incurred to acquire one user through advertising, AI-based performance reward calculation device including
3. In Article 1, The above normalization function is An AI-based performance reward calculation device that corrects the units and distributions of each KPI item in the above KPI data and converts them so that they can be compared based on a common standard.
4. In Article 1, The above reward score calculation unit An AI-based performance reward calculation device that calculates the above deduction score by summing a predetermined basic deduction value for each violation type and a predetermined AI correction value from a predetermined policy violation log received as input.
5. In Article 4, The above reward score calculation unit An AI-based performance reward calculation device that calculates the AI correction value by weighted averaging the correction prediction value calculated from the XGBoost model and the non-linear prediction value calculated from the Multi-Layer Perceptron (MLP) model.
6. In Article 5, The above reward score calculation unit An AI-based performance reward calculation device that adjusts the weights applied to the XGBoost model and the weights applied to the MLP model based on the temporal trend of the above KPI data.
7. In Article 1, The above reward score calculation unit An AI-based performance reward calculation device that calculates the risk score by applying a predetermined monotonically decreasing function to the probability value of the possibility of abnormal behavior of the evaluation subject output using a predetermined anomaly detection model based on the above KPI data.
8. In Article 7, The above monotonically decreasing function is An AI-based performance reward calculation device corresponding to a logarithmic function in which the output value decreases as the probability value of the above-mentioned abnormal behavior increases.
9. In Article 1, The above reward score calculation unit An AI-based performance reward calculation device that calculates the sustainability score by weighting the predicted values calculated using a first time series analysis model that analyzes a predetermined short-term trend in the KPI data based on the above past performance data and a second time series analysis model that analyzes a predetermined long-term trend.
10. In Article 1, The above differential compensation amount calculation unit An AI-based performance reward calculation device that maximizes the correlation between the above-mentioned reward amount and the performance indicator in the above-mentioned KPI data, and learns the above-mentioned global weights based on an objective function that considers a predetermined fairness constraint value.
11. In Article 10, The above differential compensation amount calculation unit An AI-based performance reward calculation device that calculates the contribution of certain violation items in the above KPI data and calculates the above fairness constraint value to minimize bias among violation items.
12. In Article 1, The above differential compensation amount calculation unit An AI-based performance reward calculation device that calculates a reward amount by applying a predetermined non-linear reward function having a different slope such that the reward amount has a different rate of change as the reward score increases, for each predetermined reward score interval corresponding to each of the multiple evaluation subjects.
13. In Article 12, The above non-linear compensation function is An AI-based performance reward calculation device corresponding to at least one of a sigmoid shape configured to continuously change the rate of change of the reward amount for the above reward score and a piecewise linear shape composed of a plurality of linear segments.
14. In Article 13, The above non-linear compensation function is An AI-based performance reward calculation device having a reward cap and floor applied so that the above reward amount does not exceed a predetermined upper or lower limit.
15. In the AI-based performance reward calculation method of an AI-based performance reward calculation device, A step of receiving specific Key Performance Indicator (KPI) data for multiple evaluation subjects; A compensation step of calculating a predetermined performance-based score, deduction score, risk score, and sustainability score for each of the plurality of evaluation subjects by applying a weight based on the contribution of each indicator learned from past performance data through an AI learning model to the above KPI data; and A step of calculating differential reward amounts for each of the plurality of evaluation subjects based on reward scores calculated by applying predetermined global weights learned through an AI learning model to each of the scores calculated for each of the plurality of evaluation subjects; AI-based performance reward calculation method including
16. In a computer program stored on a computer-readable recording medium comprising instructions for providing an AI-based performance reward calculation method, Receiving specific Key Performance Indicator (KPI) data for multiple evaluation subjects, and By applying weights based on the contribution of each indicator learned from past performance data through an AI learning model and a predetermined normalization function to the above KPI data, predetermined performance-based scores, deduction scores, risk scores, and sustainability scores are calculated for each of the above multiple evaluation subjects. A computer program stored on a computer-readable recording medium, comprising a sequence of instructions for calculating differential reward amounts for each of the plurality of evaluation subjects based on a reward score calculated by applying predetermined global weights learned through an AI learning model to each of the scores calculated for each of the plurality of evaluation subjects.

Description

Apparatus, Method and Program for Calculating Performance Reward Based on Artificial Intelligence The present invention relates to an AI-based performance reward calculation device, method, and program. In the online advertising and digital marketing industries, the integrated management of data by campaign, platform, and target is becoming increasingly complex due to the proliferation of multi-channel environments. Existing marketing management systems evaluated performance based on a limited number of quantitative indicators, such as clicks, conversion rates, and revenue from individual advertising channels; however, this approach had limitations in that it failed to adequately reflect correlations between channels or external factors (seasonality, events, competitive advertising, etc.). In particular, when evaluating the performance of advertisers and marketers, qualitative factors (content quality, policy violations, user responses, etc.) were not properly reflected, resulting in a discrepancy between actual efficiency and evaluation scores. Consequently, fairness in the performance compensation process was compromised, and problems existed where the actual scale of damage or the risk of recurrence was not sufficiently considered due to the uniform application of penalty points for identical violations. On the other hand, since large-scale platforms must simultaneously evaluate numerous marketers and campaign data, it is difficult to ensure both evaluation precision and processing speed using only manual reviews or simple rule-based penalty methods. Furthermore, if the regulatory violation detection process relies on human judgment, there is a possibility of disputes arising from subjectivity or a lack of consistency. Recently, attempts have been made to utilize artificial intelligence (AI) to predict advertising efficiency and automate budget execution; however, these technologies primarily focus on prediction and operational automation, leaving them insufficient for correcting violations or implementing performance compensation systems that reflect risk levels. Therefore, there is a need for a technological approach that can satisfy automation, accuracy, and fairness in marketing performance evaluation and compensation calculation. FIG. 1 is a configuration diagram for explaining an AI-based performance reward calculation system according to one embodiment of the present invention. FIG. 2 is a configuration diagram for explaining an AI-based performance reward calculation device according to one embodiment of the present invention. FIG. 3 is a flowchart illustrating an AI-based performance reward calculation method according to an embodiment of the present invention. Embodiments of the present invention are described below with reference to the attached drawings so that those skilled in the art can easily implement the invention. However, the present invention may be embodied in various different forms and is not limited to the embodiments described herein. Furthermore, in order to clearly explain the present invention in the drawings, parts unrelated to the explanation have been omitted, and similar parts throughout the specification are denoted by similar reference numerals. Throughout the specification, when a part is described as being "connected" to another part, this includes not only cases where they are "directly connected" but also cases where they are "electrically connected" with other elements interposed between them. Furthermore, when a part is described as "including" a component, this means that, unless specifically stated otherwise, it does not exclude other components but rather allows for the inclusion of additional components; it should be understood that this does not preclude the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof. In this specification, the term "part" includes a unit realized by hardware, a unit realized by software, and a unit realized using both. Additionally, one unit may be realized using two or more hardware, and two or more units may be realized by one hardware. Some of the operations or functions described in this specification as being performed by a terminal or device may instead be performed by a server connected to said terminal or device. Likewise, some of the operations or functions described as being performed by a server may also be performed by a terminal or device connected to said server. An embodiment of the present invention will be described in detail below with reference to the attached drawings. FIG. 1 is a configuration diagram for explaining an AI-based performance reward calculation system according to an embodiment of the present invention. Referring to FIG. 1, the AI-based performance reward calculation system (1) may include an AI-based performance reward calculation device (100) and a user terminal (200). The AI-based performance reward calculation device (100) and