Shopping Terminal Prompting With Sensor-to-Label Conversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Large language models are not effective in handling numerical values from sensors, such as temperature and humidity, leading to inadequate responses to environmental conditions.
Innovation Solution
A shopping terminal that converts sensor data into qualitative label text using a classification system, incorporating this text into prompts for a large language model to generate contextually relevant answers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If sensor data is directly input to a large language model, then the model can process the data, but the model cannot accurately reflect environmental conditions in its responses
Solution Approach 1:
The patent introduces an intermediary component that converts numerical sensor data into qualitative label text before inputting it to the large language model. This mediator bridges the gap between the model's text-processing strengths and the numerical nature of sensor data, enabling accurate environmental condition reflection without requiring the model to directly handle numerical values it was not designed to process effectively.
2Ease of operation
If numerical sensor data is used directly, then data precision is maintained, but the large language model cannot generate contextually appropriate responses
Solution Approach 1:
The system transforms the parameter representation of sensor data from numerical values to qualitative labels (e.g., converting temperature values like 25°C to labels like 'warm'). This parameter transformation enables the large language model to process the data effectively while still capturing the essential environmental conditions needed for generating contextually appropriate recommendations.
3Reliability
If sensor data is converted to label text, then the model can generate appropriate responses, but numerical information is lost
Solution Approach 1:
The system extracts the essential qualitative characteristics from numerical sensor data by converting them into label text. Rather than attempting to preserve all numerical information, the extraction process identifies and retains only the critical environmental condition attributes (such as temperature range, humidity level) that are necessary for generating contextually relevant responses, discarding redundant numerical precision that the language model cannot effectively utilize.
Data Source
AI summary
A shopping terminal used in a store includes an input device, an interface circuit connectable to sensors located outside or inside the store, a display, a memory, and a processor configured to execute a program stored in the memory to perform: acquiring sensor data representative of environmental conditions from the sensors through the interface circuit, acquiring first text that is input through the input device, converting each of the sensor data into second text, generating a prompt using the first and second text, inputting the prompt to a computer model, which generates in response thereto third text that promotes an item sold in the store, the computer model being a large language model that has learned relationships and connections between human perceptions under different environmental conditions, and data of items sold in the store, and controlling the display to display the third text.