Sentiment Analysis for Predicting Pharmaceutical Prescribing Likelihood

Solution Overview

Problem

Enterprises face challenges in efficiently analyzing large-scale client and potential client data to predict events and optimize supply chain operations, particularly in the pharmaceutical industry where digital meetings have reduced the effectiveness of prescribing pharmaceutical compositions, necessitating an automated method to analyze digital conversation transcripts and sentiment analysis.

Innovation Solution

A computer-implemented method using a trained machine learning model to derive sentiment scores from transcripts and combine them with engagement data to predict the likelihood of actions, such as prescribing pharmaceuticals, by classifying emotional content and determining notification thresholds for automated actions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual analysis of interaction records is used, then accuracy of event prediction can be maintained, but productivity and reaction time deteriorate due to the large volume of data

Engineering Contradiction:

Core Design Contradiction:

Measurement precisionVSProductivity

Solution Approach 1:

The patent replaces manual mechanical analysis of interaction records with automated machine learning models and natural language processing systems. These computational systems process large volumes of transcript and engagement data rapidly, maintaining prediction accuracy while dramatically improving productivity and reaction time.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent introduces machine learning models and sentiment analysis algorithms as intermediary systems between raw interaction data and business decisions. These intermediaries automatically process and interpret unstructured data, extracting meaningful insights without requiring manual review while preserving prediction accuracy.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Loss of time

If digital meetings are used to reduce meeting frequency, then loss of time is reduced, but reliability of prescribing behavior prediction deteriorates due to shorter meeting durations

Engineering Contradiction:

Core Design Contradiction:

Loss of timeVSReliability

Solution Approach 1:

The patent compensates for reduced meeting duration by adding multiple additional data dimensions beyond transcript content. These include digital engagement data, sentiment scores, interaction patterns, and historical behavior data. By analyzing across these multiple dimensions, the system maintains reliable prescribing behavior predictions despite shorter meeting times.

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

Solution Approach 2:

The patent performs preliminary analysis of multiple data sources before and during meetings to build comprehensive predictive models. By pre-processing and analyzing engagement data, sentiment trends, and historical interactions, the system accumulates sufficient information for accurate predictions even when actual meeting duration is limited.

Inventive Principle:

Principle #10Preliminary action

3Productivity

If automated machine learning analysis is implemented, then productivity and reaction time are improved, but device complexity increases due to multiple data sources and processing steps

Engineering Contradiction:

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The patent employs multi-functional machine learning models that simultaneously process transcripts, sentiment analysis, engagement data, and historical records through a unified system architecture. This universal approach consolidates multiple processing functions into integrated models, managing system complexity while maintaining high productivity and comprehensive analysis capabilities.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Data Source

PatentUS20240370657A1Machine learning methods to determine a likelihood for an event to occur through sentiment analysis of digital conversations

Publication Date: 2024.11.07

AI summary

A computer-implemented method can comprise accessing a trained learning machine, evaluating, using the machine learning model, the transcript to output a first sentiment score related to the first party in the unique domain, accessing digital engagement data representing engagement of the first party with digital assets associated with the second party, evaluating the one or more sentiment score values and the digital engagement data to output a value indicative of a likelihood of the first party to take a particular action, and determining whether the value is above a threshold, and if so, automatically sending a notification to a computer device associated with the second party.