skip to main content

Title: Evaluation of Seven Part-of-Speech Taggers in Tagging Building Codes: Identifying the Best Performing Tagger and Common Sources of Errors
As the number, size and complexity of building construction projects increase, code compliance checking becomes more challenging because of the time-consuming, costly, and error-prone nature of a manual checking process. A fully automated code compliance checking would be desirable in facilitating a more efficient, cost effective, and human error-proof code checking. Such automation requires automated information extraction from building designs and building codes, and automated information transformation to a format that allows automated reasoning. Natural Language Processing (NLP) is an important technology to support such automated processing of building codes, because building codes are represented in natural language texts. Part-of-speech (POS) tagging, as an important basis of NLP tasks, must have a high performance to ensure the quality of the automated processing of building codes in such a compliance checking system. However, no systematic testing of existing POS taggers on domain specific building codes data have been performed. To address this gap, the authors analyzed the performance of seven state-of-the-at POS taggers on tagging building codes and compared their results to a manually-labeled gold standard. The authors aim to: (1) find the best performing tagger in terms of accuracy, and (2) identify common sources of errors. In providing the POS more » tags, the authors used the Penn Treebank tagset, which is a widely used tagset with a proper balance between conciseness and information richness. An average accuracy of 88.80% was found on the testing data. The Standford coreNLP tagger outperformed the other taggers in the experiment. Common sources of errors were identified to be: (1) word ambiguity, (2) rare words, and (3) unique meaning of common English words in the construction context. The found result of machine taggers on building codes calls for performance improvement, such as error-fixing transformational rules and machine taggers that are trained on building codes. « less
Award ID(s):
Publication Date:
Journal Name:
ASCE Construction Research Congress 2020
Sponsoring Org:
National Science Foundation
More Like this
  1. Traditional manual building code compliance checking is costly, time-consuming, and human error-prone. With the adoption of Building Information Modeling (BIM), automation in such a checking process becomes more feasible. However, existing methods still face limited automation when applied to different building codes. To address that, in this paper, the authors proposed a new framework that requires minimal input from users and strives for full automation, namely, the Invariant signature, logic reasoning, and Semantic Natural language processing (NLP)-based Automated building Code compliance Checking (I-SNACC) framework. The authors developed an automated building code compliance checking (ACC) prototype system under this framework and tested it on Chapter 10 of the International Building Codes 2015 (IBC 2015). The system was tested on two real projects and achieved 95.2% precision and 100% recall in non-compliance detection. The experiment showed that the framework is promising in building code compliance checking. Compared to the state-of-the-art methods, the new framework increases the degree of automation and saves manual efforts for finding non-compliance cases.
  2. Steinmetz, A. (Ed.)
    Manual building code compliance checking is a time-consuming, labor-intensive and error-prone process. Automated logic-based reasoning is an essential step in the automation of this process. There have been previous studies using logic programming languages for automated logic-based reasoning to support automated compliance checking (ACC) of building designs with building codes. As a high-performance implementation of the standard logic programming language, B-Prolog was widely used in these studies. However, due to the support of dynamic predicates and user-defined operators, the predicates’ functions vary according to different user definitions; therefore, B-Prolog is sometimes not reliable for building code reasoning. As a more expressive, scalable, and reliable alterative to B-Prolog, Picat, a logic-based multi-paradigm programming language, provides a new and potentially more powerful platform for automated logic-based reasoning in ACC. To explore the potential value of Picat in ACC, in this study, the authors compared Picat and B-Prolog performance in automatically checking 20 requirement rules in the 2015 International Building Code. The experimental results showed that the automated checking for building codes in the B-Prolog version was faster than that in the Picat version, whereas the Picat version was more reliable than the B-Prolog version. This could be the result of B-Prolog usingmore »unifica-tion and Picat using pattern matching for indexing rules. More potential applications of Picat in ACC domain need further research. Furthermore, this schema could be used in the teaching of ACC to graduate construction students, illustrating the need to focus on the reliability, predictability and scalability of the process, in order to provide a practical solution to improving code compliance checking processes.« less
  3. To allow full automation of building code compliance checking with different building design models and codes/regulations, input building design models need to be automatically validated. Automated architecture, engineering, and construction (AEC) object identification with high accuracy is essential for such validation. For example, in order to check egress requirements, exits of a building (and their presence or absence) need to be identified automatically through object identification. To address that, the authors propose a new AEC object identification algorithm that can identify needed code checking concepts from building design models based on the invariant signatures of AEC objects, which consisted of Cartesian points-based geometry, relative location and orientation, and material mechanical properties. Building design models in industry foundation classes (IFC) format are processed into invariant signatures, which can fully represent the model data and convert them into computable representations to support automated compliance reasoning. A systematic implementation of the above invariant signatures-based object identification algorithm can be used to automatically conduct building design model validation for code compliance checking preparation. An experimental testing on Chapters 4 and 8 of the International Building Code 2015 and a convenience store design model showed the model validation using the proposed identification algorithms successfully validatedmore »ceiling and interior door concepts. Comparing to the manual validation used in current practice, this new object identification algorithm is more efficient in supporting model validation for automated building code compliance checking.« less
  4. Part-of-speech (POS) tagging is the foundation of many natural language processing applications. Rule-based POS tagging is a wellknown solution, which assigns tags to the words using a set of predefined rules. Many researchers favor statistical-based approaches over rule-based methods for better empirical accuracy. However, until now, the computational cost of rule-based POS tagging has made it difficult to study whether more complex rules or larger rulesets could lead to accuracy competitive with statistical approaches. In this paper, we leverage two hardware accelerators, the Automata Processor (AP) and Field Programmable Gate Arrays (FPGA), to accelerate rule-based POS tagging by converting rules to regular expressions and exploiting the highly-parallel regular-expressionmatching ability of these accelerators. We study the relationship between rule set size and accuracy, and observe that adding more rules only poses minimal overhead on the AP and FPGA. This allows a substantial increase in the number and complexity of rules, leading to accuracy improvement. Our experiments on Treebank and Brown corpora achieve up to 2,600X and 1,914X speedups on the AP and on the FPGA respectively over rule-based methods on the CPU in the rule-matching stage, up to 58× speedup over the Perceptron POS tagger on the CPU in total testingmore »time, and up to 253× speedup over the LSTM tagger on the GPU in total testing time, while showing a competitive accuracy compared to neural-network and statistical solutions.« less
  5. Unsupervised cross-lingual projection for part-of-speech (POS) tagging relies on the use of parallel data to project POS tags from a source language for which a POS tagger is available onto a target language across word-level alignments. The projected tags then form the basis for learning a POS model for the target language. However, languages with rich morphology often yield sparse word alignments because words corresponding to the same citation form do not align well. We hypothesize that for morphologically complex languages, it is more efficient to use the stem rather than the word as the core unit of abstraction. Our contributions are: 1) we propose an unsupervised stem-based cross-lingual approach for POS tagging for low-resource languages of rich morphology; 2) we further investigate morpheme-level alignment and projection; and 3) we examine whether the use of linguistic priors for morphological segmentation improves POS tagging. We conduct experiments using six source languages and eight morphologically complex target languages of diverse typologies. Our results show that the stem-based approach improves the POS models for all the target languages, with an average relative error reduction of 10.3% in accuracy per target language, and outperforms the word-based approach that operates on three-times more data formore »about two thirds of the language pairs we consider. Moreover, we show that morpheme-level alignment and projection and the use of linguistic priors for morphological segmentation further improve POS tagging.« less