skip to main content

Explore Research Products in the NSF-PAR

Title: ASPER: Answer Set Programming Enhanced Neural Network Models for Joint Entity-Relation Extraction
Abstract A plethora of approaches have been proposed for joint entity-relation (ER) extraction. Most of these methods largely depend on a large amount of manually annotated training data. However, manual data annotation is time-consuming, labor-intensive, and error-prone. Human beings learn using both data (through induction) and knowledge (through deduction). Answer Set Programming (ASP) has been a widely utilized approach for knowledge representation and reasoning that is elaboration tolerant and adept at reasoning with incomplete information. This paper proposes a new approach, ASP-enhanced Entity-Relation extraction (ASPER), to jointly recognize entities and relations by learning from both data and domain knowledge. In particular, ASPER takes advantage of the factual knowledge (represented as facts in ASP) and derived knowledge (represented as rules in ASP) in the learning process of neural network models. We have conducted experiments on two real datasets and compare our method with three baselines. The results show that our ASPER model consistently outperforms the baselines.  more » « less
Award ID(s):
1914635
NSF-PAR ID:
10443252
Author(s) / Creator(s):
; ;
Date Published:
Journal Name:
Theory and Practice of Logic Programming
ISSN:
1471-0684
Page Range / eLocation ID:
1 to 17
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. ( , Springer)
    Proc. 2023 European Conf. on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (Ed.)
    Automated relation extraction without extensive human-annotated data is a crucial yet challenging task in text mining. Existing studies typically use lexical patterns to label a small set of high-precision relation triples and then employ distributional methods to enhance detection recall. This precision-first approach works well for common relation types but struggles with unconventional and infrequent ones. In this work, we propose a recall-first approach that first leverages high-recall patterns (e.g., a per:siblings relation normally requires both the head and tail entities in the person type) to provide initial candidate relation triples with weak labels and then clusters these candidate relation triples in a latent spherical space to extract high-quality weak supervisions. Specifically, we present a novel framework, RCLUS, where each relation triple is represented by its head/tail entity type and the shortest dependency path between the entity mentions. RCLUS first applies high-recall patterns to narrow down each relation type’s candidate space. Then, it embeds candidate relation triples in a latent space and conducts spherical clustering to further filter out noisy candidates and identify high-quality weakly-labeled triples. Finally, RCLUS leverages the above-obtained triples to prompt-tune a pre-trained language model and utilizes it for improved extraction coverage. We conduct extensive experiments on three public datasets and demonstrate that RCLUS outperforms the weakly-supervised baselines by a large margin and achieves generally better performance than fully-supervised methods in low-resource settings. 
    more » « less
  2. ; ; ; ; ; ; ; ; ; ( , IEEE)
    2023 4th International Conference on Big Data Analytics and Practices (Ed.)
    Many publicly available datasets exist that can provide factual answers to a wide range of questions that benefit the public. Indeed, datasets created by governmental and nongovernmental organizations often have a mandate to share data with the public. However, these datasets are often underutilized by knowledge workers due to the cumbersome amount of expertise and embedded implicit information needed for everyday users to access, analyze, and utilize their information. To seek solutions to this problem, this paper discusses the design of an automated process for generating questions that provide insight into a dataset. Given a relational dataset, our prototype system architecture follows a five-step process from data extraction, cleaning, pre-processing, entity recognition using deep learning, and questions formulation. Through examples of our results, we show that the questions generated by our approach are similar and, in some cases, more accurate than the ones generated by an AI engine like ChatGPT, whose question outputs while more fluent, are often not true to the facts represented in the original data. We discuss key limitations of our approach and the work to be done to bring to life a fully generalized pipeline that can take any data set and automatically provide the user with factual questions that the data can answer. 
    more » « less
  3. ; ; ; ; ( , Proceedings of the 30th ACM International Conference on Information & Knowledge Management)
    Knowledge graphs (KGs) are of great importance in various artificial intelligence systems, such as question answering, relation extraction, and recommendation. Nevertheless, most real-world KGs are highly incomplete, with many missing relations between entities. To discover new triples (i.e., head entity, relation, tail entity), many KG completion algorithms have been proposed in recent years. However, a vast majority of existing studies often require a large number of training triples for each relation, which contradicts the fact that the frequency distribution of relations in KGs often follows a long tail distribution, meaning a majority of relations have only very few triples. Meanwhile, since most existing large-scale KGs are constructed automatically by extracting information from crowd-sourcing data using heuristic algorithms, plenty of errors could be inevitably incorporated due to the lack of human verification, which greatly reduces the performance for KG completion. To tackle the aforementioned issues, in this paper, we study a novel problem of error-aware few-shot KG completion and present a principled KG completion framework REFORM. Specifically, we formulate the problem under the few-shot learning framework, and our goal is to accumulate meta-knowledge across different meta-tasks and generalize the accumulated knowledge to the meta-test task for error-aware few-shot KG completion. To address the associated challenges resulting from insufficient training samples and inevitable errors, we propose three essential modules neighbor encoder, cross-relation aggregation, and error mitigation in each meta-task. Extensive experiments on three widely used KG datasets demonstrate the superiority of the proposed framework REFORM over competitive baseline methods. 
    more » « less
  4. ; ; ( , Proceedings of the 28th Conference on Nuclear Engineering)
    null (Ed.)
    Abstract

    In the Nuclear Power Plant (NPP) control room, the operators’ performance in emergencies is impacted by the need to monitor many indicators on the control room boards, the limited time to interact with dynamic events, and the incompleteness of the operator’s knowledge. Recent research has been directed toward increasing the level of automation in the NPP system by employing modern AI techniques that support the operator’s decisions. In previous work, the authors have employed a novel AI-guided declarative approach (namely, Answer Set Programming (ASP)) to represent and reason with human qualitative knowledge. This represented knowledge is structured to form a reasoning-based operator support system that assists the operator and compensates for any knowledge incompleteness by performing reasoning to diagnose failures and recommend executing actions in real time. A general ASP code structure has been proposed and tested against simple scenarios, e.g., diagnosis of pump failures that result in loss of flow transients and generating the needed plans for resolving the issue of stuck valves in the secondary loop.

    In this work, we investigate the potential of the previously proposed ASP structure by applying ASP to a realistic case study of the Three Mile Island, Unit 2 (TMI-2) accident event sequence (in particular, the first 142 minutes). The TMI scenario presents many challenges for a reasoning system, including a large number of variables, the complexity of the scenario, and the misleading readings. The capability of the ASP-based reasoning system is tested for diagnosis and recommending actions throughout the scenario. This paper is the first work to test and demonstrate the capability of an automated reasoning system by applying it to a realistic nuclear accident scenario, such as the TMI-2 accident.

     
    more » « less
  5. ; ; ; ( , Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing)
    Humans subconsciously engage in geospatial reasoning when reading articles. We recognize place names and their spatial relations in text and mentally associate them with their physical locations on Earth. Although pretrained language models can mimic this cognitive process using linguistic context, they do not utilize valuable geospatial information in large, widely available geographical databases, e.g., OpenStreetMap. This paper introduces GeoLM, a geospatially grounded language model that enhances the understanding of geo-entities in natural language. GeoLM leverages geo-entity mentions as anchors to connect linguistic information in text corpora with geospatial information extracted from geographical databases. GeoLM connects the two types of context through contrastive learning and masked language modeling. It also incorporates a spatial coordinate embedding mechanism to encode distance and direction relations to capture geospatial context. In the experiment, we demonstrate that GeoLM exhibits promising capabilities in supporting toponym recognition, toponym linking, relation extraction, and geo-entity typing, which bridge the gap between natural language processing and geospatial sciences. The code is publicly available at https://github.com/knowledge-computing/geolm. 
    more » « less
  • Citation Formats
  • MLA
  • APA
  • Chicago
  • BibTeX
  • Save / Share this Record
  • Send to Email