Email Query Spelling Correction via Contextual Ranking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Email search systems lack effective spelling correction capabilities due to sparse and private user data, unlike web search which benefits from extensive session logs and a universal document collection, leading to decreased user productivity from repeated query reformulations.
Innovation Solution
A ranker-based algorithm that generates and ranks candidate corrections for misspelled email queries using personalized email data features, including language, translation, content-based, and contextual features, to provide accurate and context-specific corrections.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If machine learning strategies are used for spelling correction, then spelling correction accuracy is improved, but data availability is worsened due to sparse and private email data
Solution Approach 1:
The patent introduces an intermediary approach by combining web-based spelling correction (leveraging external data) with email-specific contextual features (leveraging local data). The system first generates candidates using web-based correction, then filters and ranks them using email-specific features like recipient names, subject lines, and body content, thus bridging the gap between data scarcity and correction accuracy
Solution Approach 2:
The patent applies local quality by tailoring the spelling correction process to the specific characteristics of email data. Instead of using a uniform correction approach, the system adapts the candidate filtering and ranking processes to leverage email-specific features such as contact names, subject lines, and message bodies, making the correction locally optimized for email contexts
2Measurement precision
If web-based spelling correction is used, then general spelling accuracy is improved, but email-specific context relevance is worsened
Solution Approach 1:
The patent applies local quality by tailoring the spelling correction process to the specific characteristics of email data. Instead of using a uniform correction approach, the system adapts the candidate filtering and ranking processes to leverage email-specific features such as contact names, subject lines, and message bodies, making the correction locally optimized for email contexts
Solution Approach 2:
The system uses feedback from email-specific features to refine and rank the spelling correction candidates. By evaluating candidates against actual email content (recipient names, subject lines, body text), the system receives feedback that helps select the most relevant corrections for the specific email context
Data Source
AI summary
Techniques and constructs to facilitate spelling correction of email queries can leverage features of email data to obtain candidate corrections particular to the email data being queried. The constructs may enable accurate spelling correction of email queries across languages and domains based on, for example, one or more of a language model such as a bigram language model and/or a normalized token IDF based language model, a translation model such as an edit distance translation model and/or a fuzzy match translation model, content-based features, and/or contextual features. Content-based features can include features associated with the subject line of emails, content including identified phrases, contacts, and/or the number of candidate emails returned. Contextual features can include a time window of subject match and/or contact match, a frequency of emails received from a contact, and/or device characteristics.