A SYSTEM THAT ENABLES INTELLIGENT PRODUCT RANKING.

Abstract

This invention relates to a system (1) developed to enable personalized smart product ranking and to ensure that the right product is presented to the customer at the right time by analyzing navigation data such as brand, category, gender, views, clicks, add to cart, purchase, favorites, discount rate, color, stock, availability, season, year and profitability using machine learning algorithms.

Description

2024.490 1 TARIFF A SYSTEM THAT ENABLES INTELLIGENT PRODUCT RANKING. Technical Area This invention uses machine learning algorithms to analyze by brand, category, gender, and display. Navigation data such as click count, add to cart count, purchase count, and favorite count, Analysis of information such as discount rate, color, stock, availability, season, year, and profitability. This allows for personalized smart product rankings and ensures that the right product is delivered to the right place. It is related to a system that ensures the customer is connected with the system in a timely manner. 10 Previous Technique In e-commerce, product ranking is crucial for connecting the right product with the right customer. It is of critical importance. However, incorrect sequencing or failure to meet customer demands... The prominence given to unsuitable products negatively impacts sales and... It damages the customer experience. Products at the right time and in the right categories. Not listing it could cause potential buyers to lose interest or miss the product they are looking for. This leads to them being unable to find what they are looking for. This, in turn, results in customer dissatisfaction and high basket abandonment rates. This can lead to lower rates and ultimately low conversion rates. E-commerce 20 platforms, product through user data, shopping history and behavioral analysis Optimizing their rankings will both increase sales and improve customer loyalty. It is of great importance to ensure this. Therefore, machine learning algorithms are used to analyze brand, category, gender, views, 25 Navigation data such as click count, add to cart count, purchase count, and favorite count, Analysis of information such as discount rate, color, stock, availability, season, year, and profitability. This allows for personalized smart product rankings and ensures that the right product is delivered to the right place. 2024.490 2 A system was needed that would ensure the product and the customer were brought together in a timely manner. It is understood. According to Chinese patent document number CN115293859, which is included in the prior art, AI-based e-commerce platform product smart recommendation management system 5 It is mentioned that the invention in question is an e-commerce platform based on artificial intelligence. The platform describes the product intelligent recommendation management system, the technical aspects of the computer application. It is related to the field and is an AI-based e-commerce platform product smart The suggestion management method is applied to the steps of the target user. The goal is to acquire core information and obtain a similar user base through transition 10. Don't do that; it's a target for past online shopping records and numerous similar users. Obtaining numerous past online shopping records, numerous suggested products obtaining the group, obtaining numerous similarity indices and similarity index obtaining a descending list, reverse matching the suggested product group with the descending list, much Scanning and targeting 15 through an intelligent scanning model to obtain a number of scan results. To create a suggested product set and push the target suggested product set to the target user. It includes the steps. Brief Description of the Invention The aim of this invention is to use machine learning algorithms to identify brands, categories, genders, numbers such as views, clicks, add to cart, purchases, and favorites. browsing data, discount rate, color, stock, availability, season, year, and profitability. Analyzing the data to create a personalized smart product ranking and making the right choices. 25 developed to ensure the product reaches the customer at the right time The goal is to implement the system. 2024.490 3 Another aim of this invention is to improve the e-commerce customer experience and customer a system developed to increase customer satisfaction to accomplish. Another aim of this invention is to promote environmental awareness and sustainable 5 a system developed to encourage consumption to accomplish. Another aim of this invention is to contribute to economic and social impacts, Creating new jobs and workplaces through the listing of sustainable and ethical products, saving energy and 10 In order to have a positive impact on factors that reduce carbon costs It is about implementing a developed system. Detailed Description of the Invention The "Smart Product Sorting" system was developed to achieve the purpose of this invention. A "System Providing" is shown in the attached figure; Figure 1 shows a schematic view of the system that is the subject of the invention. The parts shown in the figure are individually numbered, and the corresponding numbers correspond to these numbers. It is given below. 1. System 2. Database 25 3. Server 2024.490 4 Machine learning algorithms analyze data based on brand, category, gender, views, and clicks. Navigation data such as add to cart, purchase, and favorite count, discounts by analyzing information such as rate, color, stock, availability, season, year and profitability Creating a personalized smart product list and delivering the right product at the right time. The system in question, developed to ensure that the invention is brought together with the customer (1); 5 - product browsing history, click counts, sales information, profitability information, etc. information such as when it was last sold, sales frequency, location, and stock levels. at least one database (2) structured to ensure that records are kept. - product characteristics and their relationships with each other using statistical methods examine, correlation test, independent samples t-test, ANOVA (Analysis of Variance), 10 Performing chi-square relationship tests, numerical data with uneven distribution To detect anomalies in variables, to identify existing variables in retail. Obtaining frequently used new metrics, and identifying newly defined products more Because there was no prior navigation data, making these products more visible, new Product navigation data shows the 15 most similar products within their category. populating with navigation data, extracting new variables from existing variables, machine learning in the form of purchase probability, revenue increase probability and profitability score using algorithms to obtain a product's purchaseability score, integration tests. to perform, conduct function and performance tests and change user conducting performance tests to respond to behavioral needs, test 20 Necessary revisions and additions will be made in line with the results obtained from the stages. to enable improvements and personalized smart product ranking It includes at least one server (3) configured to enable this. The database (2) in the system (1) which is the subject of the invention, can use any communication protocol 25 to communicate with the server (3) and exchange data using It is structured. The database (2) is managed by the server (3). It is being structured. 2024.490 The server (3) in the system (1) which is the subject of the invention, uses any communication protocol to communicate with the database (2) and exchange data using The server (3) is configured to manage the database (2). is configured. Server (3) stores the past browsing data of the products on mobile and / or 5 from the big data environment as a clickstream over the web It is configured to enable the collection of sales information, profitability. Server (3) Information affecting sales includes when it was last sold, sales frequency, navigation, and stock levels. It is structured to ensure that the required information is collected from relevant sources. Server (3) statistical techniques for the relationships between features 10 It is structured to enable examination with the server (3), the distribution is proper log transformation and box-cox transformation for non-existent variables to enable the application of anomaly detection (outlier detection) techniques is configured. The server (3) displays, clicks, and navigation data from the product. New rates are obtained for adding to cart and purchasing, resulting in multiple linear 15 It is structured to help manage connectivity problems. Server (3), K-Nearest Neighbor Algorithm (KNN) and similarity algorithms By using navigation data for new products, they are organized within their own categories. to ensure that the most similar products are populated with browsing data It is structured. The server (3) uses a machine learning algorithm to create 20 products. To classify the probability of purchase as 0-1 based on turnover data. It is structured. Server (3) uses a logistic regression model with which explanatory Determining the effect of the variable on whether it is purchased or not. The model should produce a mathematical function output, and each explanatory variable should be... Determining the extent to which the independent variable has an effect and assigning weights of 25. to enable the use of odds values ​​of explanatory variables It is configured. Server (3) is for products whose turnover is greater than zero (0). using regression algorithms, optimal set of variable selection methods 2024.490 6 determining the analytical hierarchical ranking of coefficients obtained from regression. to ensure its use and to determine sales growth points for each product It is configured. Server (3) displays the last 120 sales including the last sales date, frequency and profitability. RFM (monetary value) scores are obtained using daily sales data. This involves calculating the added value of each RFM segment and assigning a profitability score of 5. To ensure that products are determined by weighting them according to their rising values. It is structured in such a way. The server (3) is being developed through various testing and development activities. integration tests are to be carried out, and these tests are to be conducted in three stages. the implementation, including the performance of standard integration tests in the first stage, In the second stage, the relevant company's team will assess the function and performance within the company. conducting the tests, creating a data test within those tests, and evaluating the suitability of the data, different procedures to be performed on the dataset Testing all functions and performance targets with test scenarios and In the third and final stage, testing is carried out as a beta version on the company's infrastructure. It is configured to enable its implementation. Server (3), demographic 15 features, sales information, profitability information, when it was last sold, sales frequency, Features in the form of navigation data are data in the form of Google Big Query. It is configured to enable collection through the platform. Server (3), Generating personalized smart product ranking suggestions and sending them to relevant users. It is configured to enable transmission via the interface. 20 Industrial Applicability Thanks to the system (1) which is the subject of the invention, brand, category, machine learning algorithms can be used to identify brands, categories, Gender, views, clicks, add to cart, purchases, favorites count: 25 Navigation data such as discount rate, color, stock, availability, season, year, and profitability. Personalized smart product rankings are created by analyzing information in this format. And this ensures that the right product is delivered to the customer at the right time. 2024.490 7 Based on these fundamental concepts, the invention focuses on "Intelligent Product Sequencing". It is possible to develop a wide variety of applications related to "A System Providing (1)". and the invention cannot be limited to the examples described here, but mainly to the claims as stated.

Claims

2024.490 8 REQUESTS 1. Machine learning algorithms for brand, category, gender, display, Navigation patterns include clicks, add to cart, purchases, and favorited items. data in 5 categories: discount rate, color, stock, availability, season, year, and profitability. By analyzing the data, a personalized smart product ranking is created and... ensuring the right product is delivered to the customer at the right time; - product browsing history, click counts, sales information, profitability Information such as when it was last sold, sales frequency, location, and stock levels. EN 10 is structured to ensure that information in this form is kept on record. containing a small database (2) and - product characteristics and their relationships with each other using statistical methods examine, correlation test, independent samples t-test, ANOVA (Analysis of Variance) (Analysis), performing chi-square correlation tests, uniform distribution to detect anomalies in non-existent numerical variables, existing 15 Obtaining new metrics commonly used in retail from variables, new Since the products identified as such did not have prior browsing data making these products more visible, navigation data for new products, our own navigation data of the most similar products within its categories fill, obtain new variables from existing variables, purchase 20 machine learning in the form of probability, probability of revenue increase and profitability score Using algorithms to obtain a product's purchaseability score, integration. to conduct tests, to perform function and performance tests, and Performance tests to respond to changing user behavior needs. to carry out, based on the results obtained from the testing phases, the necessary 25 to ensure revisions and additional improvements are made and personalized at least one configured to enable smart product ranking. a system characterized by server (3) (1). 2024.490 9 2. To communicate with the server (3) using any communication protocol and with the database (2) configured to carry out data exchange A system like the one in Claim 1, characterized (1).

3. Database (2) configured to be managed by the server (3) A system like the one in Claim 1 or 2, characterized (1).

4. Communicate with the database (2) using any communication protocol. and characterized by the server (3) configured to carry out data exchange 10 a system like any of the above-mentioned requests (1).

5. Characterized by the server (3) configured to manage the database (2). a system like any of the above-mentioned requests (1).

6. Click-through data for product browsing history via mobile and / or web. the collection of big data from the big data environment as a stream (clickstream) the above characterized by the server (3) configured to provide a system like any of the requests (1).

7. Sales information, profitability information, when it was last sold, sales frequency, Navigation, information affecting sales in the form of inventory from relevant sources. characterized by the server (3) configured to enable the collection a system like any of the above requests (1).

8. Determining the relationships between features using statistical techniques. characterized by the server (3) configured to enable its examination a system like any of the above requests (1). 2024.490 9. Log transformation and box-cox for irregularly distributed variables Transformation-based anomaly detection (outlier detection) techniques characterized by the server (3) configured to enable its implementation a system like any of the above requests (1).

10. Product navigation data including views, clicks, add to cart, and purchases. Multicollinearity problems by obtaining new ratios in the manner of acquisition. characterized by the server (3) configured to enable the handling of the problem. a system like any of the above-mentioned requests (1).

11. K-Nearest Neighbor Algorithm (KNN) and similarity algorithms By using navigation data for new products within their own categories to ensure that browsing data from the most similar products is used to populate the system. from the above requests characterized by the server (3) configured for a system like any other (1). 15 12. Using a machine learning algorithm to determine purchases based on product revenue data. structured to classify the probability of receiving as 0-1 as in any of the above requests characterized by the server (3) a system (1). 20 13. Which explanatory variable is purchased and / or which is purchased using a logistic regression model? the determination of the effect on the non-receivable class, the mathematical analysis of the model the function should output and each descriptive variable should be independent. Determining the extent to which the variable has an effect and assigning weights of 25. To enable the use of odds values ​​of explanatory variables. from the above requests characterized by the server (3) configured for a system like any other (1). 2024.490 11 14. A regression algorithm for products with turnover greater than zero (0) the use of variable selection methods to determine the optimal set, The coefficients obtained from the regression are arranged in an analytical hierarchical order. to ensure its use and to determine sales growth points for each product. 5 of the above requests characterized by the server (3) configured to do so. a system like any other (1).

15. Data in the form of sales over the last 120 days, including the last sales date, frequency, and profitability. using RFM (monetary value) points to obtain each RFM Calculating the added value of the segment and the profitability score of the products. to ensure that it is determined by weighting according to the rising values any of the above requests characterized by the configured server (3) a system like one of them (1).

16. Conducting various testing and development studies, as well as integration tests, 15 The tests in question should be carried out in three stages, the first stage being standard integration tests will be carried out, in the second phase by the relevant company's team. to carry out function and performance tests within its organization, These tests involve creating a data test and assessing the suitability of the data. evaluation, different test scenarios to be performed on the dataset 20 testing all functions and performance targets and third-party In the final stage, testing will be carried out as a beta version on the company's infrastructure. characterized by the server (3) configured to enable its implementation. a system like any of the above-mentioned requests (1).

17. Demographic characteristics, sales data, profitability information, when it was last sold, Features such as sales frequency and navigation data are used in Google Big Query. to ensure that it is collected through a data platform in the form of 2024.490 12 any of the above requests characterized by the configured server (3) a system like one of them (1).

18. Generating personalized smart product ranking suggestions and related... 5 configured to ensure that it is delivered to users via an interface as in any of the above requests characterized by the server (3) a system (1).

Images