Figure 1 - available via license: Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International
Content may be subject to copyright.
Download
View publication
Confusion matrices of each category for ChatGPT with the chain-of-thought (CoT) approach. The Rest class indicates predictions on all other classes.

Confusion matrices of each category for ChatGPT with the chain-of-thought (CoT) approach. The Rest class indicates predictions on all other classes.

Source publication
Figure 1: Confusion matrices of each category for ChatGPT with the...
Figure 2: Correlation between sample length and frequency of...
Analyzing the results, it becomes evident that ChatGPT faces difficulty...
Characterizing Information Seeking Events in Health-Related Social Discourse
Preprint
Full-text available
  • Aug 2023
Social media sites have become a popular platform for individuals to seek and share health information. Despite the progress in natural language processing for social media mining, a gap remains in analyzing health-related texts on social discourse in the context of events. Event-driven analysis can offer insights into different facets of healthcar...
Cite
Download full-text

Context in source publication

Context 1
... to go down to 1.25mg pretty soon too. Figure 1 illustrates the confusion matrices for the Chat-GPT chain-of-thought approach. Table 5 presents the classwise overprediction ratio (#false positive / #pre-dicted positives) for both ChatGPT and XLNet. ...