SPEAR : Semi-supervised Data Programming in Python

TL;DR

SPEAR is an open-source Python library that simplifies semi-supervised data programming by enabling rule-based labeling, noisy label aggregation, and training for text classification, integrating multiple approaches for weak supervision.

Contribution

The paper introduces SPEAR, a comprehensive Python package that combines recent data programming techniques, including cascade and joint approaches, with user-defined labeling functions for weak supervision.

Findings

Implemented multiple label aggregation methods.

Facilitated weak supervision for text classification.

Provided extensive documentation and real-world use cases.

Abstract

We present SPEAR, an open-source python library for data programming with semi supervision. The package implements several recent data programming approaches including facility to programmatically label and build training data. SPEAR facilitates weak supervision in the form of heuristics (or rules) and association of noisy labels to the training dataset. These noisy labels are aggregated to assign labels to the unlabeled data for downstream tasks. We have implemented several label aggregation approaches that aggregate the noisy labels and then train using the noisily labeled set in a cascaded manner. Our implementation also includes other approaches that jointly aggregate and train the model for text classification tasks. Thus, in our python package, we integrate several cascade and joint data-programming approaches while also providing the facility of data programming by letting the user define labeling functions or rules. The code and tutorial notebooks are available at https://github.com/decile-team/spear. Further, extensive documentation can be found at https://spear-decile.readthedocs.io/. Video tutorials demonstrating the usage of our package are available here. We also present some real-world use cases of SPEAR.