Automated Neural Patent Landscaping in the Small Data Regime

TL;DR

This paper introduces an automated neural system for patent landscaping that performs well with minimal labeled data, using a novel data collection method, and provides a new dataset for AI patents.

Contribution

The paper presents a neural patent landscaping approach that achieves high accuracy with few labeled examples and introduces a new high-quality dataset using active learning.

Findings

Improved $F_1$ score of 0.69 on hard examples

Achieves 0.75 $F_1$ with as few as 24 training examples

Develops a new data collection procedure combining seed/anti-seed and active learning

Abstract

Patent landscaping is the process of identifying all patents related to a particular technological area, and is important for assessing various aspects of the intellectual property context. Traditionally, constructing patent landscapes is intensely laborious and expensive, and the rapid expansion of patenting activity in recent decades has driven an increasing need for efficient and effective automated patent landscaping approaches. In particular, it is critical that we be able to construct patent landscapes using a minimal number of labeled examples, as labeling patents for a narrow technology area requires highly specialized (and hence expensive) technical knowledge. We present an automated neural patent landscaping system that demonstrates significantly improved performance on difficult examples (0.69 $F_1$ on 'hard' examples, versus 0.6 for previously reported systems), and also significant improvements with much less training data (overall 0.75 $F_1$ on as few as 24 examples). Furthermore, in evaluating such automated landscaping systems, acquiring good data is challenge; we demonstrate a higher-quality training data generation procedure by merging Abood and Feltenberger's (2018) "seed/anti-seed" approach with active learning to collect difficult labeled examples near the decision boundary. Using this procedure we created a new dataset of labeled AI patents for training and testing. As in prior work we compare our approach with a number of baseline systems, and we release our code and data for others to build upon.