Human-in-the-loop systems for cleaning NLP training data rely on automated sieves to isolate potentially-incorrect labels for manual review. We have developed a novel technique for flagging potentially-incorrect labels with high sensitivity in named entity recognition corpora. We incorporated our sieve into an end-to-end system for cleaning NLP corpora, implemented as a modular collection of Jupyter notebooks built on extensions to the Pandas DataFrame library. We used this system to identify incorrect labels in the CoNLL-2003 corpus for English-language named entity recognition (NER), one of the most influential corpora for NER model research. Unlike previous work that only looked at a subset of the corpus’s validation fold, our automated sieve enabled us to examine the entire corpus in depth. Across the entire CoNLL-2003 corpus, we identified over 1300 incorrect labels (out of 35089 in the corpus). We have published our corrections, along with the code we used in our experiments. We are developing a repeatable version of the process we used on the CoNLL-2003 corpus as an open-source library.

The CoNLL-2003 corpus for English-language named entity recognition (NER) is one of the most influential corpora for NER model research. A large number of publications, including many landmark works, have used this corpus as a source of ground truth for NER tasks. In this paper, we examine this corpus and identify over 1300 incorrect labels (out of 35089 in the corpus). In particular, the number of incorrect labels in the test fold is comparable to the number of errors that state-of-the-art models make when running inference over this corpus. We describe the process by which we identified these incorrect labels, using novel variants of techniques from semi-supervised learning. We also summarize the types of errors that we found, and we revisit several recent results in NER in light of the corrected data. Finally, we show experimentally that our corrections to the corpus have a positive impact on three state-of-the-art models.

The rise of enterprise applications over unstructured and semi-structured documents poses new challenges to text understanding systems across multiple dimensions. We present SystemT, a declarative text understanding system that addresses these challenges and has been deployed in a wide range of enterprise applications. We highlight the design considerations and decisions behind SystemT in addressing the needs of the enterprise setting. We also summarize the impact of SystemT on business and education.

The rise of Big Data analytics over unstructured text has led to renewed interest in information extraction (IE). These applications need effective IE as a first step towards solving end-to-end real world problems (e.g. biology, medicine, finance, media and entertainment, etc). Much recent NLP research has focused on addressing specific IE problems using a pipeline of multiple machine learning techniques. This approach requires an analyst with the expertise to answer questions such as: “What ML techniques should I combine to solve this problem?”; “What features will be useful for the composite pipeline?”; and “Why is my model giving the wrong answer on this document?”. The need for this expertise creates problems in real world applications. It is very difficult in practice to find an analyst who both understands the real world problem and has deep knowledge of applied machine learning. As a result, the real impact by current IE research does not match up to the abundant opportunities available.In this tutorial, we introduce the concept of transparent machine learning. A transparent ML technique is one that:- produces models that a typical real world use can read and understand;- uses algorithms that a typical real world user can understand; and- allows a real world user to adapt models to new domains.The tutorial is aimed at IE researchers in both the academic and industry communities who are interested in developing and applying transparent ML.