This work presents a new task-aware prompt design and example retrieval approach for information extraction (IE) using a prompt chaining technique. Our approach divides IE tasks into two steps: (1) text classification to understand what information (e.g., entity or event types) is contained in the underlying text and (2) information extraction for the identified types. Initially, we use a large language model (LLM) in a few-shot setting to classify the contained information. The classification output is used to select the relevant prompt and retrieve the examples relevant to the input text. Finally, we ask a LLM to do the information extraction with the generated prompt. By evaluating our approach on legal IE tasks with two different LLMs, we demonstrate that the prompt chaining technique improves the LLM’s overall performance in a few-shot setting when compared to the baseline in which examples from all possible classes are included in the prompt. Our approach can be used in a low-resource setting as it does not require a large amount of training data. Also, it can be easily adapted to many different IE tasks by simply adjusting the prompts. Lastly, it provides a cost benefit by reducing the number of tokens in the prompt.

This work presents a manually annotated dataset for Information Extraction (IE) from legal wills, and relevant in-context learning experiments on the dataset. The dataset consists of entities, binary relations between the entities (e.g., relations between testator and beneficiary), and n-ary events (e.g., bequest) extracted from 45 legal wills from two US states. This dataset can serve as a foundation for downstream tasks in the legal domain. Another use case of this dataset is evaluating the performance of large language models (LLMs) on this IE task. We evaluated GPT-4 with our dataset to investigate its ability to extract information from legal wills. Our evaluation result demonstrates that the model is capable of handling the task reasonably well. When given instructions and examples as a prompt, GPT-4 shows decent performance for both entity extraction and relation extraction tasks. Nevertheless, the evaluation result also reveals that the model is not perfect. We observed inconsistent outputs (given a prompt) as well as prompt over-generalization.

This study investigates whether a legal natural language inference (NLI) model trained on the data from one US state can be transferred to another state. We fine-tuned a pre-trained model on the task of evaluating the validity of legal will statements, once with the dataset containing the Tennessee wills and once with the dataset containing the Idaho wills. Each model’s performance on the in-domain setting and the out-of-domain setting are compared to see if the models can across the states. We found that the model trained on one US state can be mostly transferred to another state. However, it is clear that the model’s performance drops in the out-of-domain setting. The F1 scores of the Tennessee model and the Idaho model are 96.41 and 92.03 when predicting the data from the same state, but they drop to 66.32 and 81.60 when predicting the data from another state. Subsequent error analysis revealed that there are two major sources of errors. First, the model fails to recognize equivalent laws across states when there are stylistic differences between laws. Second, difference in statutory section numbering system between the states makes it difficult for the model to locate laws relevant to the cases being predicted on. This analysis provides insights on how the future NLI system can be improved. Also, our findings offer empirical support to legal experts advocating the standardization of legal documents.

This work introduces a natural language inference (NLI) dataset that focuses on the validity of statements in legal wills. This dataset is unique because: (a) each entailment decision requires three inputs: the statement from the will, the law, and the conditions that hold at the time of the testator’s death; and (b) the included texts are longer than the ones in current NLI datasets. We trained eight neural NLI models in this dataset. All the models achieve more than 80% macro F1 and accuracy, which indicates that neural approaches can handle this task reasonably well. However, group accuracy, a stricter evaluation measure that is calculated with a group of positive and negative examples generated from the same statement as a unit, is in mid 80s at best, which suggests that the models’ understanding of the task remains superficial. Further ablative analyses and explanation experiments indicate that all three text segments are used for prediction, but some decisions rely on semantically irrelevant tokens. This indicates that overfitting on these longer texts likely happens, and that additional research is required for this task to be solved.