Natural Language Processing (NLP) Models

To analyze off-chain textual data, such as customer feedback or social media activity, Trumnix integrates state-of-the-art NLP models.

Key Capabilities:

  1. Sentiment Analysis: Detects sentiment in user reviews or community discussions to refine borrower profiling.

  2. Entity Recognition: Identifies key borrower attributes (e.g., employer names, locations) in text data.

  3. Topic Modeling: Categorizes textual data into relevant topics for better decision-making.

Protocols:

  1. Data Cleaning Protocol: Removes noise (e.g., typos, spam) from input text to improve NLP accuracy.

  2. Contextual Analysis Protocol: Applies transformers (e.g., BERT, GPT) for contextual understanding of textual data.

Last updated