Emotional Arc Waveform Generation for Narrative Text Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies lack efficient methods for automatically identifying and characterizing emotion arcs within narrative text, which is crucial for evaluating narrative quality.
Innovation Solution
The approach involves tokenizing narrative text, extracting events using multiple techniques, and generating a waveform to characterize emotional arcs. These waveform elements are then used to score the narrative quality using a machine learning model.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If automated text analysis methods are used to evaluate narrative quality, then scoring efficiency is improved, but the ability to capture emotional arcs and narrative depth deteriorates
Solution Approach 1:
The narrative text is segmented into discrete events using event extraction techniques. Each event is analyzed for its emotional valence and contribution to the overall narrative arc, allowing the system to process large texts efficiently while capturing nuanced emotional changes at the event level.
Solution Approach 2:
An emotional arc waveform is introduced as an intermediary representation that bridges the gap between automated text processing and emotional narrative analysis. The waveform transforms discrete event data into a continuous emotional trajectory, enabling both efficient computation and precise emotional characterization.
2Measurement precision
If multiple extraction techniques are used in parallel to identify events, then event detection completeness is improved, but system complexity increases
Solution Approach 1:
Multiple extraction techniques (regex-based extraction, lexicon-based extraction, and machine learning-based extraction) are merged into a unified parallel processing framework. The techniques operate simultaneously on the tokenized text, with their results aggregated to produce comprehensive event detection without requiring complex sequential processing.
3Stability of the object's composition
If waveform smoothing techniques are applied to emotional arcs, then noise reduction is improved, but emotional detail preservation deteriorates
Solution Approach 1:
The smoothing process is applied partially rather than excessively. The system applies smoothing to reduce high-frequency noise in the emotional waveform while preserving the major emotional arcs and significant emotional transitions. This balanced approach maintains waveform stability without over-smoothing important emotional details.
Data Source
AI summary
Quality of a narrative is characterized by receiving data that includes a narrative text. This narrative text is then tokenized and events are extracted from the tokenized words. The extraction can use, in parallel, two or more different extraction techniques. The extracted events are then extracted so that a waveform can be generated based on the aggregated extracted events that characterizes a plurality of emotional arcs within the narrative text. Subsequently, a plurality of waveform elements are extracted from the waveform. The narrative quality (or other quality) of the narrative text is then scored based on the extracted plurality of waveform elements and using a machine learning model trained to correlate emotional arc waveforms with narrative quality scores. Related apparatus, systems, techniques and articles are also described.