Message Prioritization Using Actionable Request Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The abundance of low-priority messages overwhelms important requests and questions in email traffic, leading to frustration and delays in responding to critical correspondence.
Innovation Solution
An apparatus and method that analyze message content to identify actionable questions and requests, predict response times, and categorize replies, using AI to filter out non-essential messages and alert users to pending responses, allowing for timely action on important requests.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If all messages are treated equally in the inbox, then no message is prioritized, but important requests are buried and response time increases
Solution Approach 1:
The patent applies local quality by treating different messages differently based on their characteristics. It analyzes message content to identify questions, directives, and requests, then applies different prioritization rules to different message types. Important messages from preferred senders are surfaced to the top while routine messages remain in the standard queue, creating localized quality variations in message handling.
Solution Approach 2:
The system performs preliminary action by pre-analyzing incoming messages to identify their type and importance before the user sees them. It automatically detects questions, directives, and requests in advance, determines sender preferences, and pre-sorts messages into prioritized categories before presentation, saving the user time and effort.
2Productivity
If the system analyzes all message content to identify important requests, then response prioritization improves, but processing complexity and computational resources increase
Solution Approach 1:
The patent extracts only the essential elements from messages that indicate importance - specifically identifying questions, directives, and requests through keyword detection and pattern matching. Rather than analyzing entire message contents deeply, it extracts key indicators such as interrogative words, imperative phrases, and specific syntax patterns, reducing computational overhead while maintaining effectiveness.
Solution Approach 2:
The system changes parameters by focusing analysis on specific linguistic features rather than comprehensive content analysis. It detects particular word patterns, sentence structures, and grammatical indicators that signal actionable items, transforming the analysis task from holistic to targeted, thereby improving efficiency.
3Loss of time
If the system surfaces only high-priority messages, then response time to important requests improves, but low-priority messages may be overlooked
Solution Approach 1:
The patent segments the message queue into distinct priority levels and categories. Instead of creating a single prioritized list that might hide lower-priority items, it organizes messages into multiple segments - urgent questions, important directives, routine requests, and standard correspondence. This segmentation ensures that while high-priority messages are surfaced first, lower-priority messages remain visible in their own categories and are not completely overlooked.
Data Source
AI summary
When a message contains an actionable request or question the invention measures a time til transmitted reply (TTTR). Replies which contain substantive answers such as dates, decisions, quantities, concurrence, or content are paired with their respective request for each sender-recipient directed dyad. The replies are categorized to the type of action taken in response to the request. The history of each dyad is examined to determine the relative importance each party ascribes to the other and a range of typical TTTR is stored. Questions, directives, and requests are discovered in a message by scanning for deadline words, calendar words, time words, urgency phrases, question words such as who what when, capability syntax, and agreement/disagreement alternatives. A user receives an alert when a response is nearly past a deadline or close to exceeding expected time to reply, immediately upon receipt, or both. An alert offers a potential response.