Alert button

Large-Scale Text Analysis Using Generative Language Models: A Case Study in Discovering Public Value Expressions in AI Patents

May 18, 2023
Sergio Pelaez, Gaurav Verma, Barbara Ribeiro, Philip Shapira

Share this with someone who'll enjoy it:

Labeling data is essential for training text classifiers but is often difficult to accomplish accurately, especially for complex and abstract concepts. Seeking an improved method, this paper employs a novel approach using a generative language model (GPT-4) to produce labels and rationales for large-scale text analysis. We apply this approach to the task of discovering public value expressions in US AI patents. We collect a database comprising 154,934 patent documents using an advanced Boolean query submitted to InnovationQ+. The results are merged with full patent text from the USPTO, resulting in 5.4 million sentences. We design a framework for identifying and labeling public value expressions in these AI patent sentences. A prompt for GPT-4 is developed which includes definitions, guidelines, examples, and rationales for text classification. We evaluate the quality of the labels and rationales produced by GPT-4 using BLEU scores and topic modeling and find that they are accurate, diverse, and faithful. These rationales also serve as a chain-of-thought for the model, a transparent mechanism for human verification, and support for human annotators to overcome cognitive limitations. We conclude that GPT-4 achieved a high-level of recognition of public value theory from our framework, which it also uses to discover unseen public value expressions. We use the labels produced by GPT-4 to train BERT-based classifiers and predict sentences on the entire database, achieving high F1 scores for the 3-class (0.85) and 2-class classification (0.91) tasks. We discuss the implications of our approach for conducting large-scale text analyses with complex and abstract concepts and suggest that, with careful framework design and interactive human oversight, generative language models can offer significant advantages in quality and in reduced time and costs for producing labels and rationales.

* 13 pages, 3 figures. Update corrects ORCID ID, adds arXiv citation. No other changes  
View paper onarxiv icon

Share this with someone who'll enjoy it: