An official website of the United States government
Physical and digital documents often contain visually rich information. With such information, there is no strict order- ing or positioning in the document where the data values must appear. Along with textual cues, these documents often also rely on salient visual features to define distinct semantic boundaries and augment the information they disseminate. When performing information extraction (IE), traditional techniques fall short, as they use a text-only representation and do not consider the visual cues inherent to the layout of these documents. We propose VS2, a generalized approach for information extraction from heterogeneous visually rich documents. There are two major contributions of this work. First, we propose a robust segmentation algorithm that de- composes a visually rich document into a bag of visually iso- lated but semantically coherent areas, called logical blocks. Document type agnostic low-level visual and semantic fea- tures are used in this process. Our second contribution is a distantly supervised search-and-select method for identify- ing the named entities within these documents by utilizing the context boundaries defined by these logical blocks. Ex- perimental results on three heterogeneous datasets suggest that the proposed approach significantly outperforms its text-only counterparts on all datasets. Comparing it against the state-of-the-art methods also reveal that VS2 performs comparably or better on all datasets.
more »
« less
Engineering of an artificial intelligence safety data sheet document processing system for environmental, health, and safety compliance
Chemical Safety Data Sheets (SDS) are the primary method by which chemical manufacturers communicate the ingredients and hazards of their products to the public. These SDSs are used for a wide variety of purposes ranging from environmental calculations to occupational health assessments to emergency response measures. Although a few companies have provided direct digital data transfer platforms using xml or equivalent schemata, the vast majority of chemical ingredient and hazard communication to product users still occurs through the use of millions of PDF documents that are largely loaded through manual data entry into downstream user databases. This research focuses on the reverse engineering of SDS document types to adapt to various layouts and the harnessing of meta-algorithmic and neural network approaches to provide a means of moving industrial institutions towards a digital universal SDS processing methodology. The complexities of SDS documents including the lack of format standardization, text and image combinations, and multi-lingual translation needs, combined, limit the accuracy and precision of optical character recognition tools. The approach in this document is to translate entire SDSs from thousands of chemical vendors, each with distinct formatting, to machine-encoded text with a high degree of accuracy and precision. Then the system will "read" and assess these documents as a human would; that is, ensuring that the documents are compliant, determining whether chemical formulations have changed, ensuring reported values are within expected thresholds, and comparing them to similar products for more environmentally friendly alternatives.
more »
« less
- Award ID(s):
- 1842577
- PAR ID:
- 10289105
- Date Published:
- Journal Name:
- DocEng '21: Proceedings of the 21st ACM Symposium on Document Engineering
- Volume:
- 21
- Issue:
- 1
- Page Range / eLocation ID:
- 1 to 4
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
null (Ed.)Text categorization is an essential task in Web content analysis. Considering the ever-evolving Web data and new emerging categories, instead of the laborious supervised setting, in this paper, we focus on the minimally-supervised setting that aims to categorize documents effectively, with a couple of seed documents annotated per category. We recognize that texts collected from the Web are often structure-rich, i.e., accompanied by various metadata. One can easily organize the corpus into a text-rich network, joining raw text documents with document attributes, high-quality phrases, label surface names as nodes, and their associations as edges. Such a network provides a holistic view of the corpus’ heterogeneous data sources and enables a joint optimization for network-based analysis and deep textual model training. We therefore propose a novel framework for minimally supervised categorization by learning from the text-rich network. Specifically, we jointly train two modules with different inductive biases – a text analysis module for text understanding and a network learning module for class-discriminative, scalable network learning. Each module generates pseudo training labels from the unlabeled document set, and both modules mutually enhance each other by co-training using pooled pseudo labels. We test our model on two real-world datasets. On the challenging e-commerce product categorization dataset with 683 categories, our experiments show that given only three seed documents per category, our framework can achieve an accuracy of about 92%, significantly outperforming all compared methods; our accuracy is only less than 2% away from the supervised BERT model trained on about 50K labeled documents.more » « less
-
Abstract The goal of this paper is to describe the science verification of Milky Way Mapper (MWM) APOGEE Stellar Parameter and Chemical Abundances Pipeline (ASPCAP) data products published in Data Release 19 (DR19) of the fifth phase of the Sloan Digital Sky Survey (SDSS-V). We compare MWM ASPCAP atmospheric parametersTeff, logg, 24 abundances of 21 elements (carbon, nitrogen, and oxygen have multiple sources for deriving their abundance values) and their uncertainties determined from Apache Point Observatory Galactic Evolution Experiment (APOGEE) spectrograph spectra with those of the literature and evaluate their accuracy and precision. We also test the zero-point calibration of thevradderived by the APOGEE Data Reduction Pipeline. This data release contains ASPCAP parameters for 964,989 stars, including all APOGEE-2 targets expanded with new observations of 336,511 stars from the Apache Point Observatory observed until 2023 July 4. Overall, the newTeffvalues show excellent agreement with the IRFM scale, while the surface gravities exhibit slight systematic offsets compared to asteroseisimic gravities. The estimated precision ofTeffis between 50 and 70 K for giants and 70–100 K for dwarfs, while surface gravities are measured with a precision of 0.07–0.09 dex for giants. We achieve an estimated precision of 0.02–0.04 dex for multiple elements, including metallicity,α, Mg, and Si, while the precision of at least 10 elements is better than 0.1 dex.more » « less
-
Automating the annotation of scanned documents is challenging, requiring a balance between computational efficiency and accuracy. DocParseNet addresses this by combining deep learning and multi-modal learning to process both text and visual data. This model goes beyond traditional OCR and semantic segmentation, capturing the interplay between text and images to preserve contextual nuances in complex document structures. Our evaluations show that DocParseNet significantly outperforms conventional models, achieving mIoU scores of 49.12 on validation and 49.78 on the test set. This reflects a 58% accuracy improvement over state-of-the-art baseline models and an 18% gain compared to the UNext baseline. Remarkably, DocParseNet achieves these results with only 2.8 million parameters, reducing the model size by approximately 25 times and speeding up training by 5 times compared to other models. These metrics, coupled with a computational efficiency of 0.039 TFLOPs (BS=1), highlight DocParseNet's high performance in document annotation. The model's adaptability and scalability make it well-suited for real-world corporate document processing applications.more » « less
-
null (Ed.)Categorizing documents into a given label hierarchy is intuitively appealing due to the ubiquity of hierarchical topic structures in massive text corpora. Although related studies have achieved satisfying performance in fully supervised hierarchical document classification, they usually require massive human-annotated training data and only utilize text information. However, in many domains, (1) annotations are quite expensive where very few training samples can be acquired; (2) documents are accompanied by metadata information. Hence, this paper studies how to integrate the label hierarchy, metadata, and text signals for document categorization under weak supervision. We develop HiMeCat, an embedding-based generative framework for our task. Specifically, we propose a novel joint representation learning module that allows simultaneous modeling of category dependencies, metadata information and textual semantics, and we introduce a data augmentation module that hierarchically synthesizes training documents to complement the original, small-scale training set. Our experiments demonstrate a consistent improvement of HiMeCat over competitive baselines and validate the contribution of our representation learning and data augmentation modules.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- https://doi.org/10.1145/3469096.3474933
