CN-121998226-A - System and method for generating content summaries with minimal illusions

CN121998226ACN 121998226 ACN121998226 ACN 121998226ACN-121998226-A

Abstract

Systems and methods for generating a content summary with minimal illusions are provided. In particular, the systems and methods may be applicable to EV charging stations. One or more natural language processing models (NLP models), such as an aspect-based emotion analysis model (ABSA) and a summary model, may be used to generate a summary of information associated with the charging station. A mechanism may also be used to ensure that the generated summary is accurate and does not include an illusion.

Inventors

V. Reba
N. Sadatijafakalei
WU KAI
Pu Jiongzhu

Assignees

福特全球技术公司

Dates

Publication Date: 20260508
Application Date: 20251021
Priority Date: 20241028

Claims (15)

1. A system, comprising: A memory storing computer executable instructions, and One or more processors configured to access the memory and execute the computer-executable instructions to: Receiving first input data from one or more data sources at a first time, the first input data including a first user rating of one or more Electric Vehicle (EV) charging stations; Causing a first model to predict first output data, the first output data comprising a first category associated with the first input data; Causing a second model to generate one or more first summaries based on the first input data and the first output data; causing a third model to determine that the one or more first summaries include an illusion created by the second model; Causing the second model to generate one or more second summaries based on determining that the one or more first summaries include the illusions generated by the second model, and The one or more second summaries are presented via a user interface of the device.
2. The system of claim 1, wherein the first model is an aspect-based emotion analysis (ABSA) model and the second model is a summary model.
3. The system of claim 1, wherein determining the one or more first summaries comprises an illusion further comprising: causing the third model to generate one or more first questions using the first input data and the one or more first summaries; Causing the third model to generate one or more first answers to the one or more first questions based on the first input data and one or more second answers to the one or more first questions based on the one or more first summaries, and A difference between the one or more first answers and the one or more second answers is determined.
4. The system of claim 3, wherein determining the one or more first summaries comprises an illusion further comprising: causing the third model to refine the one or more first questions to produce one or more second questions; Causing the third model to generate one or more third answers to the one or more second questions based on the first input data and one or more second answers to the one or more second questions based on the one or more first summaries, and Determining a second difference between the one or more first answers and the one or more second answers, wherein causing the second model to generate the one or more second summaries is based on the second difference.
5. The system of claim 1, wherein the one or more processors are further configured to execute the computer-executable instructions to: The first input data is preprocessed to at least one of remove empty user ratings or modify a format of user ratings in the user ratings.
6. The system of claim 1, wherein the one or more processors are further configured to execute the computer-executable instructions to: Receiving second input data from the one or more data sources at a second time, the second input data including second user ratings of one or more Electric Vehicle (EV) charging stations; determining that the second input data includes a threshold number of user ratings, and Based on determining that the second input data includes a threshold number of user ratings, the second model is caused to update the one or more second summaries based on the first input data and the first output data.
7. The system of claim 1, wherein the first category comprises at least one of functionality and reliability, accessibility and availability, location, price, convenience, user tips, charging speed and efficiency, customer service, or charging station security, wherein the first output data further comprises emotion of user assessment in the first user assessment, the emotion comprising at least one of positive, negative, or neutral.
8.A method, comprising: Receiving first input data from one or more data sources at a first time, the first input data including a first user rating of one or more Electric Vehicle (EV) charging stations; Causing a first model to predict first output data, the first output data comprising a first category associated with the first input data; Causing a second model to generate one or more first summaries based on the first input data and the first output data; causing a third model to determine that the one or more first summaries include an illusion created by the second model; Causing the second model to generate one or more second summaries based on determining that the one or more first summaries include the illusions generated by the second model, and The one or more second summaries are presented via a user interface of the device.
9. The method of claim 8, wherein the first model is an aspect-based emotion analysis (ABSA) model and the second model is a summary model.
10. The method of claim 8, further comprising: causing the third model to generate one or more first questions using the first input data and the one or more first summaries; Causing the third model to generate one or more first answers to the one or more first questions based on the first input data and one or more second answers to the one or more first questions based on the one or more first summaries, and A difference between the one or more first answers and the one or more second answers is determined.
11. The method of claim 10, wherein determining the one or more first summaries comprises an illusion further comprising: causing the third model to refine the one or more first questions to produce one or more second questions; Causing the third model to generate one or more third answers to the one or more second questions based on the first input data and one or more second answers to the one or more second questions based on the one or more first summaries, and Determining a second difference between the one or more first answers and the one or more second answers, wherein causing the second model to generate the one or more second summaries is based on the second difference.
12. The method of claim 8, further comprising: The first input data is preprocessed to at least one of remove empty user ratings or modify a format of user ratings in the user ratings.
13. The method of claim 8, further comprising: Receiving second input data from the one or more data sources at a second time, the second input data including second user ratings of one or more Electric Vehicle (EV) charging stations; determining that the second input data includes a threshold number of user ratings, and Based on determining that the second input data includes a threshold number of user ratings, the second model is caused to update the one or more second summaries based on the first input data and the first output data.
14. The method of claim 8, wherein the first category comprises at least one of functionality and reliability, accessibility and availability, location, price, convenience, user tips, charging speed and efficiency, customer service, or charging station safety, wherein the first output data further comprises emotion of user assessment in the first user assessment, the emotion comprising at least one of positive, negative, or neutral.
15. A non-transitory computer-readable medium storing computer-executable instructions that, when executed by one or more processors, cause the one or more processors to: Receiving first input data from one or more data sources at a first time, the first input data including a first user rating of one or more Electric Vehicle (EV) charging stations; Causing a first model to predict first output data, the first output data comprising a first category associated with the first input data; Causing a second model to generate one or more first summaries based on the first input data and the first output data; causing a third model to determine that the one or more first summaries include an illusion created by the second model; Causing the second model to generate one or more second summaries based on determining that the one or more first summaries include the illusions generated by the second model, and The one or more second summaries are presented via a user interface of the device.

Description

System and method for generating content summaries with minimal illusions Technical Field The present disclosure relates to systems and methods for generating a content summary with minimal illusions. Background As more drivers rely on Electric Vehicles (EVs) for their daily transportation needs, the necessity for reliable and user-friendly information about the EV charging location becomes increasingly important. EV drivers rely heavily on customer ratings to select the appropriate charging location, but the absolute amount of ratings can make quick identification of relevant insights challenging. Currently, users must screen numerous reviews to find useful information about factors such as station functionality, charging speed, and overall experience. This process is time consuming and inefficient, especially when a fast decision is required. The different nature of customer feedback further complicates the challenge. Ratings often involve different aspects such as availability, reliability, and convenience, making it challenging for a user to obtain a comprehensive understanding from a single rating. Furthermore, the presence of both positive and negative ratings can be confusing, as users need to weigh different opinions to form an overall judgment. In addition, the public EV infrastructure of the USA encompasses various Charge Point Operators (CPOs), each providing different mobile applications, features and options. Because of the diversity of services, customers encounter unique problems at these charging stations that result in a poor charging experience. Fortunately, some previous users provided suggestions and solutions to these specific problems. However, not all customers will view these user reviews, and the process can be very time consuming for those customers viewing these user reviews. Another difficulty is that there are multiple platforms for evaluation. EV drivers often review various data sources to gather comprehensive information. This fragmented information adds complexity to making an intelligent choice because the driver spends additional time cross-referencing the assessment, which detracts from the convenience promised by EV technology. Finally, as the EV market continues to expand, the amount of evaluation will increase, exacerbating these problems. The necessity of an efficient way of synthesizing and presenting this information in a user friendly way becomes increasingly urgent. More importantly, this information may activate a digital service that assists EV drivers in identifying the optimal charging location based on their personal preferences. Disclosure of Invention The present disclosure describes systems and methods for generating a content summary with minimal illusions (e.g., incorrect information included within the summary). Although use cases are described herein to generate an overview for an EV charging station, the method may also be applicable to other use cases. A user typically desires to view a brief overview of key information including areas of interest to the user. However, such information is often not readily accessible in such concise format and requires a user to obtain multiple pieces of information from various types of data sources. This results in a time consuming and inefficient process, and in some cases, the user may not be able to locate all of the desired information. This typical process is particularly detrimental in situations where the user needs information to make a decision within a short period of time (e.g., if the user is driving and needs to identify an EV charging station to use, as described in further detail below). In the use case of EV charging stations, when a user is deciding which charging station to use to charge their vehicle (e.g., if the user is currently required to charge their vehicle and is looking for a nearby charging station, or if the user is planning a future trip and desires to pre-plan a route to include charging stations with user-desired attributes), the user may desire to view concise information regarding key aspects of the EV charging station. However, typical sources of this type of information include user ratings posted on various types of online platforms, and users may have difficulty navigating these data sources to efficiently find information about various charging stations. The systems described herein address these challenges by generating a concise and useful overview of information from various data sources. In particular, the system includes a plurality of Natural Language Processing (NLP) models. In some embodiments, the model may include an aspect-based emotion analysis model (ABSA) and an overview model, however, any other number of models (as well as combinations of different types of models) may be used. In some cases, a single model may also be used to perform all of the tasks described herein. Large Language Models (LLM) can be used to develop these models by employing both prompt eng