Personal Corpus Vector Update for Polysemous Word Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Polysemous words can create communication issues between individuals due to differences in meaning, even when the same words are present in both parties' corpora, as existing technologies lack effective methods to compare and detect variations in word meanings across personal corpora.
Innovation Solution
A system and method are developed to generate user-specific personal corpora by processing text data from various sources, including browsing, email, and chat histories, using vector-based analysis to identify and update unknown words and detect polysemous words, thereby comparing and synchronizing meanings between users.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If vector-based analysis is used to identify and update unknown words, then the accuracy of personal corpus generation is improved, but the computational complexity increases
Solution Approach 1:
The system performs preliminary actions by creating a basic corpus with initial word vectors before processing user-specific text. This pre-established foundation allows for more efficient subsequent processing of personal texts, as the system already has a structured framework to work with rather than starting from scratch during the actual personal corpus generation.
Solution Approach 2:
The patent uses an intermediary mechanism by introducing average vectors as mediators between unknown words and the existing basic corpus. When encountering unknown words, the system calculates average vectors from context words to represent these unknown terms, enabling the integration of new vocabulary without requiring direct complex processing of every unknown word against the entire corpus.
2Reliability
If the system processes and compares word meanings across multiple users' corpora, then communication accuracy is improved, but the time required for corpus synchronization increases
Solution Approach 1:
The system applies local quality by focusing comparisons and updates on specific polysemous words rather than processing the entire corpus uniformly. When a polysemous word is detected, the system locally updates the personal corpus by replacing vectors for that specific word while leaving other words unchanged, thereby reducing the overall time required for synchronization while maintaining accuracy for critical ambiguous terms.
Solution Approach 2:
The patent utilizes parameter changes by modifying vector representations dynamically based on user-specific context. The system changes the semantic parameters of words by updating their vector representations in personal corpora based on observed usage patterns, allowing efficient adaptation of word meanings over time rather than requiring complete re-synchronization of entire corpora.
3Reliability
If polysemous words are detected and flagged, then communication issues are reduced, but the complexity of corpus management increases
Solution Approach 1:
The system extracts and isolates polysemous words from the general corpus by detecting them through vector distance comparisons and flagging them separately. This extraction allows the system to manage polysemous words as a distinct category requiring special handling, while the rest of the corpus can be managed through standard procedures, thereby reducing overall management complexity through selective differentiation.
Data Source
AI summary
In an approach for generating a user-specific personal corpus, a processor creates a basic corpus for a first user using a first set of data sources, wherein the basic corpus includes one or more basic words and one or more vectors of the one or more basic words. A processor extracts a set of text from a second set of data sources associated with the first user. Responsive to finding an unknown word included in the set of text extracted, a processor updates the basic corpus, wherein the basic corpus is updated by replacing a vector of the unknown word with an average vector of the one or more basic words in the basic corpus created and registering the unknown word in a first personal corpus.