Natural Language Visualization Generation via Intent Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies require users to manually configure data attributes and graphic visualization types, making it time-consuming and costly, especially for datasets with many categorical values. Additionally, these systems struggle to handle natural language requests that do not contain predefined terms, limiting user flexibility and creativity in data exploration.
Innovation Solution
The technology employs a natural language model trained on a corpus of documents containing graphic visualizations, along with an intent classifier. This allows users to input natural language requests, which are parsed into n-grams and used to generate graphic visualizations based on discovered data attributes and determined graphic visualization types.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual configuration of data attributes and graphic visualization types is used, then precision and control of the visualization output is improved, but time consumption and operational complexity increase significantly
Solution Approach 1:
The system enables self-service by allowing users to simply provide natural language queries without needing to manually configure data attributes or visualization types. The automated system interprets the natural language, identifies relevant data attributes, selects appropriate visualization types, and generates the visualizations automatically, thus eliminating the time-consuming manual configuration process while maintaining accurate and precise visualization outputs.
Solution Approach 2:
The patent replaces the mechanical system of manual configuration and selection with an automated natural language processing system. Instead of requiring users to mechanically select data attributes and visualization types through interfaces, the system uses NLP to interpret natural language queries and automatically performs the configuration tasks, significantly reducing time consumption while preserving the precision of the output.
2Reliability
If predefined terms are required for natural language requests, then system reliability and accuracy are improved, but user flexibility and adaptability decrease
Solution Approach 1:
The system changes the parameter of language processing from requiring predefined terms to accepting flexible natural language expressions. By using advanced NLP techniques including word embeddings and intent classification, the system can accurately interpret a wide variety of user expressions without being constrained by a fixed vocabulary, thus maintaining reliability while significantly improving user flexibility and adaptability.
Solution Approach 2:
The patent implements universality by designing the system to handle multiple types of user queries and expressions through a single natural language processing framework. The intent classification and word embedding models enable the system to universally interpret different ways users might express their needs, whether using predefined terms or creative variations, thereby maintaining accuracy while enhancing versatility and user flexibility.
3Productivity
If automated natural language processing is used, then productivity and ease of operation are improved, but system complexity increases
Solution Approach 1:
The system applies preliminary action by pre-training the natural language model and intent classification system on comprehensive datasets before actual use. This pre-processing and preparation of the AI models enables the system to quickly and accurately process natural language queries during operation, significantly improving productivity without requiring complex real-time computations, thus managing system complexity while achieving high performance.
Data Source
AI summary
Graphic visualizations, such as charts or graphs conveying data attribute values, can be generated based on natural language queries, i.e., natural language requests. To do so, a natural language request is parsed into n-grams, and from the n-grams, word embeddings are determined using a natural language model. Data attributes for the graphic visualization are discovered in the vector space from the word embeddings. The type of graphic visualization can be determined based on a request intent, which is determined using a trained intent classifier. The graphic visualization is generated to include the data attribute values of the discovered data attributes, and in accordance with the graphic visualization type.