Skip to content

Latest commit

 

History

History
19 lines (15 loc) · 902 Bytes

README.md

File metadata and controls

19 lines (15 loc) · 902 Bytes
Raw

prompt-engineering-bdlt

Mini Project: Prompt Engineering GPT-3, Module: Big Data and Language Technologies, Institution: University Leipzig

Motivation

Use an existing taxonomy and extract topics and their superordinate ones.

Methods

Take topic names, insert them into following phrases and pass those as inputs into the GPT-3 text-davinci-002 model.

    'The superordinate topic for "{}" is called',
    'In mathematics, the hypernym for "{}" is called',
    'In mathematics, the superordinate topic for "{}" is called',
    'Within the mathematical taxonomy, the superordinate topic for "{}" is called',
    'Mathematical topics are strongly connected, for example the hypernym of {} is called'

Calculate a similarity score with use of Levenshtein distance. It does not matter if the superordinate topic is the next parent or an even higher level superordinate topic.