Dynamic Poster Generation System Based on LLM and Association Rule in Conversation Behavior Analysis

Abstract

The invention relates to a system that integrates AI-based natural language processing (NLP), user chat behavior analysis modeling, data mining, and automated visualization technologies to analyze thematic, behavioral, and semantic information derived from chat-based user interactions using large language models (LLM) and association rule mining methods, and transform this information into dynamic and personalized posters.

Description

1 TARIFF Dynamic Poster Based on LLM and Association Rule in Conversation Behavior Analysis Creation System Technical Area The invention utilizes AI-based natural language processing (NLP) and user conversational behavior analysis. integration of modeling, data mining and automated visualization technologies providing thematic, behavioral and 10-dimensional data derived from chat-based user interactions. Semantic information using large language models (LLM) and association rule mining methods a system that enables the analysis and transformation of images into dynamic and personalized posters It is related to. State of the Art 15 With the rapid development of digital environments today, users' interaction with digital content The methods differ; real-time analysis of user preferences, integration into content production processes and automated personalized designs This has led to expectations of its production. In current techniques, conversation analytics, conversation 20 content mining, sentiment analysis, topic modeling, association rule extraction, and LLM-based content creation. Technologies such as those used in manufacturing are widely utilized as independent components. However, these An integrated system that generates dynamic posters from chat data where technologies work together seamlessly. A systematic method is not yet available. Known poster creation methods, While this is generally possible with static data processing, visual layout is 25 based on user behavior. It does not support updating and remains limited in the production of personalized content. Similarly, association rule mining is used in retail analytics, social media behavior, and These rules are linked to the visual design engine when used in areas such as recommendation systems. No such application exists in the literature. LLM-based text generation systems, however, are mostly It produces purely textual output; it does not work in an integrated structure with visual design processes. 30 Therefore, current methods require a dynamic visual design in conversational behavior analysis. It does not offer comprehensive solutions that enable the process to be transferred. The aforementioned shortcomings, It combines user behavior analysis, association rules, and automated design into a single workflow. and offers an integrated and personalized solution that current techniques cannot provide. 35 2 Purpose of the Invention The invention was created by drawing inspiration from existing situations and overcoming the aforementioned drawbacks. It aims to solve the problem. The main purpose of the invention is to analyze user behavioral patterns derived from chat data, automatically analyzes the thematic structures and interaction intensities, and from these analyses An innovative artificial intelligence system that can generate personalized dynamic posters using the results obtained. It is about developing an intelligence system. Another aim of the invention is not just to perform textual analysis of chat data; it also... integrating relational patterns extracted from data into visual design processes to understand behavior Our goal is to provide focused visual content production. Another aim of the invention is to monitor changing user behavior over time (poster 15) The goal is to update designs in real time; used in visual design. Texts generated by LLM, icons, colors, backgrounds and composition. The aim is to ensure that association rules play a decisive role in the selection of their regulations. To achieve the objectives described above, the invention utilizes artificial intelligence-based natural language processing. (NLP), user conversational behavior analysis modeling, data mining, and automated visualization 20 chat-based user interface that enables the integration of creation technologies large language models of thematic, behavioral and semantic information obtained from their interactions Dynamic and personalized solutions are created by analyzing (LLM) and association rule mining methods. It is a system that enables the creation of posters. The structural and characteristic features and all the advantages of the invention are given in the figure below. Thanks to the detailed explanation written with references to the diagram, it becomes clearer. This will be understood, and therefore the evaluation should also take this detailed explanation into account. It must be done by taking it. Figures that will help understand the invention. Figure 1 shows the general architecture of the system that is the subject of the invention. Explanation of Part References 1. Data collection layer 35 2. User database 3 3. User database network server 4. User chat behavior analysis interface 5. User chat behavior analysis interface network server 6. Association Rule Module 7. Association rule module network server 5 8. Dynamic poster engine 9. Dynamic poster engine network server Detailed Description of the Invention In this detailed description, the preferred configurations of the system that is the subject of the invention are listed only. This is explained to facilitate a better understanding of the subject. The meanings of the terms used in this detailed explanation are as follows: The invention utilizes AI-based natural language processing (NLP) and user conversational behavior analysis. 15 integration of modeling, data mining and automated visualization technologies providing thematic, behavioral and analytical insights derived from chat-based user interactions. Semantic information using large language models (LLM) and association rule mining methods a system that enables these posters to be analyzed and transformed into dynamic and personalized posters. It is a system. 20 The working principle of the system in question is as follows: The data collection layer (1) is frequently preferred It includes interaction data of users on the chat platforms mentioned. User database network storage of user data (2) running on server (3) It provides. User chat behavior analysis interface running on network server (5) User 25 Thanks to the chat behavior analysis interface (4), sentiment analysis, keyword detection, topic Modeling and interaction intensity calculations are performed. Each resulting theme, concept, and The sentiment is converted into sets of processing data. Association rule module network server (7) Patterns detected by the association rule module (6) working on support, trust and They are evaluated according to lifting criteria, and highly reliable rules are selected for the work. 30 Dynamic poster engine visuals running on the network server (9) The layout engine automatically selects icon systems, color palettes, and design composition. It creates a poster. The system updates itself over time as new chat data arrives, and It produces constantly evolving, customized posters for the user. 35 4 The functions of the system described in the invention are as follows: Data collection: Data collection layer (1) correspondence of users on the chat platform, the interaction they have and the collection of message data by the system Data preprocessing: User database running on network server (3) (2) The area where chat user data is stored and the collected chat data is 5 cleaned, classified, and made available for analysis using natural language processing methods. coming NLP Analysis: User chat behavior analysis interface running on network server (5) User chat behavior analysis interface (4) collects the chat sessions made, cleaning and processing with NLP method 10 Association rule extraction: Transferring the obtained theme and behavior data to process sets. association rule module running on network server (7) for conversion Determining relationships between frequently occurring themes by running different algorithms with module (6) Design parameter generation: Visual design of the resulting association rules. Converting to parameters and using LLM according to the determined design parameters 15 Automatic generation of poster content texts Poster creation: Dynamic poster engine running on network server (9) Determining the relationship between the engine (8) and product, package or content preferences within the segments Dynamic updating: The behavior model is updated as new chat data arrives at the system. Calculation and poster updating 20

Claims

REQUESTS 1. Artificial intelligence-based natural language processing (NLP), user conversational behavior analysis modeling, enabling the integration of data mining and automated visualization technologies, Thematic, behavioral, and semantic analyses derived from chat-based user interactions. 5 The data is analyzed using large language models (LLM) and association rule mining methods. It is a system that enables these images to be transformed into dynamic and personalized posters. feature; • the correspondence and interactions of users on the chat platform, and their messages Data collection layer (1), which enables the collection of data by the system, 10 • the area where chat user data is stored and the collected chat data cleaning, classifying and analyzing using natural language processing methods User database (2) which enables it to become usable, • User whose server hardware has the user database (2) application installed database network server (3), 15 • Collecting, cleaning, and processing the recorded chat sessions using NLP methods User chat behavior analysis interface (4), • server on which the user chat behavior analysis interface (4) application runs User chat behavior analysis interface network server with hardware (5), • Association rule 20 that discovers relationships between frequently occurring themes in process sets module (6), • Association rule module (6) is the server hardware that runs the association rule software. module network server (7), • Dynamic poster consisting of an LLM-based text generator and visual editing engine engine (8), 25 • Dynamic poster, which is the server hardware that runs the dynamic poster engine (8) software. engine network server (9) It includes.

2. The system conforms to Claim 1, and its characteristic is that the resulting association rules are visually designed. 30 conversion to parameters and LLM according to the defined design parameters a design parameter that enables the automatic generation of poster content text using It involves production. 6 3. The system is compliant with claim 1 and its feature is; dynamic poster engine (8) and segments within Creating posters that help identify the relationship between product, packaging, or content preferences. It includes.

Images