Auto-Correction Index Segmentation for Portable Devices
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional auto-correct algorithms on portable devices often misinterpret proper nouns, product names, and unique words due to limited dictionaries, leading to irrelevant results and user frustration, as these algorithms do not account for environment-specific terms and differ across devices from various manufacturers.
Innovation Solution
The approach involves generating an environment-specific index by running auto-correction processes on unique terms and adding corrected versions as synonyms, allowing for context-based weighting and user options to select between corrected and original terms, thereby improving search results relevance across different devices and contexts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a limited dictionary is used for auto-correction on portable devices, then the device complexity is reduced and memory usage is minimized, but the accuracy of text correction deteriorates when encountering proper nouns, product names, and environment-specific terms
Solution Approach 1:
The patent segments the dictionary into multiple components: a core dictionary stored locally on the portable device and environment-specific dictionaries stored remotely on servers. The local dictionary contains commonly used words for quick auto-correction, while environment-specific dictionaries (e.g., medical terms, legal terms, product names) are stored remotely and accessed only when needed. This segmentation allows the device to maintain high text correction accuracy for specialized terms without increasing local device complexity or memory usage.
Solution Approach 2:
The patent introduces a network communication intermediary that mediates between the portable device and remote servers storing environment-specific dictionaries. When the auto-correction algorithm encounters a term not found in the local dictionary, it queries the remote server through this intermediary. The server returns the appropriate correction or confirms the original term, enabling accurate correction of proper nouns and environment-specific terms without requiring these large dictionaries to be stored locally on the device.
2Ease of operation
If auto-correction is applied to all text input on portable devices, then typing errors are corrected, but unique words and proper nouns are mistakenly corrected leading to irrelevant results and user frustration
Solution Approach 1:
The patent implements dynamic auto-correction behavior that adapts based on the input context. The system analyzes the input text to determine whether it contains proper nouns, product names, or environment-specific terminology. When such terms are detected (either through local recognition or by querying environment-specific dictionaries), the system dynamically adjusts its correction behavior to be more conservative or to seek specialized dictionary entries, rather than applying blanket auto-correction rules. This dynamic approach prevents reliable terms from being mistakenly corrected while still correcting common typing errors.
Solution Approach 2:
The patent incorporates feedback mechanisms where user corrections and rejections of auto-corrections are recorded and used to improve future correction decisions. When users manually correct auto-corrected terms or reject suggested corrections, this feedback is stored and used to adjust the auto-correction algorithm's behavior for similar terms in the future. Additionally, the system receives feedback from remote servers about the accuracy of corrections made using environment-specific dictionaries, allowing the system to learn and improve its correction reliability over time.
3Adaptability or versatility
If different manufacturers implement their own auto-correction algorithms with proprietary dictionaries, then each device can be optimized for its specific use case, but consistency and interoperability across different devices deteriorate
Solution Approach 1:
The patent creates a universal auto-correction infrastructure where environment-specific dictionaries are stored centrally on remote servers and can be accessed by any portable device regardless of manufacturer. The system defines standardized interfaces and protocols for querying these dictionaries and receiving corrections. This allows different manufacturers to optimize their devices for specific use cases (e.g., medical devices accessing medical terminology, legal devices accessing legal terminology) while maintaining cross-device consistency through the shared universal dictionary repository and standardized communication protocols.
Data Source
AI summary
Text input that is automatically “corrected” by an auto-correction process can be analyzed to determine whether to also include text as input before the correction. A set of words particular to an environment can be run through a number of auto-correct processes to determine which words are likely to be corrected, as well as the corrected versions. These “corrected” versions can be added to an index or other collection as synonyms, or alternatives, for the uncorrected words. When a request or other input is received that includes a corrected term, a determination can be made as to whether to include content for the synonymous uncorrected term. Such an approach can enable a user to obtain content that is more likely of interest to the user based at least in part upon known corrections made to terms that did not necessarily need correcting.