Modern Natural Language Processing Research Articles

Distributional hypothesis as isomorphism between word-word co-occurrence and analogical parallelograms.

Most of the modern natural language processing (NLP) techniques are based on the vector space models of language, in which each word is represented by a vector in a high dimensional space. One of the earliest successes was demonstrated by the four-term analogical reasoning task: what is to C as B is to A? The trained word vectors form "parallelograms" representing the quadruple of words in analogy. This discovery in NLP offers us insight into our understanding of human semantic representation of words via analogical reasoning. Despite successful applications of the large-scale language models, it has not been fully understood why such parallelograms emerge by learning through natural language data. As the vector space model is not optimized to form parallelograms, the key structure related to geometric shapes of word vectors is expected to be in the data, rather than the models. In the present article, we test our hypothesis that such parallelogram arrangement of word vectors readily exists in the co-occurrence statistics of language. Our approach focuses more on the data itself, and it is different from the existing theoretical approach trying to find the mechanism of parallelogram formation in the algorithms and/or vector arithmetic operations on word vectors. First, our analysis suggested that analogical reasoning is possible by decomposition of the bigram co-occurrence matrix. Second, we demonstrated the formation of a parallelepiped, a more structured geometric object than a parallelogram, by creating a small artificial corpus and its word vectors. With these results, we propose a refined form of distributional hypothesis pointing out an isomorphism between a sort of symmetry or exchangeability and word co-occurrence statistics.

Meta-learning for real-world class incremental learning: a transformer-based approach

Modern natural language processing (NLP) state-of-the-art (SoTA) deep learning (DL) models have hundreds of millions of parameters, making them extremely complex. Large datasets are required for training these models, and while pretraining has reduced this requirement, human-labelled datasets are still necessary for fine-tuning. Few-shot learning (FSL) techniques, such as meta-learning, try to train models from smaller datasets to mitigate this cost. However, the tasks used to evaluate these meta-learners frequently diverge from the problems in the real world that they are meant to resolve. This work aims to apply meta-learning to a problem that is more pertinent to the real world: class incremental learning (IL). In this scenario, after completing its training, the model learns to classify newly introduced classes. One unique quality of meta-learners is that they can generalise from a small sample size to classes that have never been seen before, which makes them especially useful for class incremental learning (IL). The method describes how to emulate class IL using proxy new classes. This method allows a meta-learner to complete the task without the need for retraining. To generate predictions, the transformer-based aggregation function in a meta-learner that modifies data from examples across all classes has been proposed. The principal contributions of the model include concurrently considering the entire support and query sets, and prioritising attention to crucial samples, such as the question, to increase the significance of its impact during inference. The outcomes demonstrate that the model surpasses prevailing benchmarks in the industry. Notably, most meta-learners demonstrate significant generalisation in the context of class IL even without specific training for this task. This paper establishes a high-performing baseline for subsequent transformer-based aggregation techniques, thereby emphasising the practical significance of meta-learners in class IL.

Machine Learning-based Human Resource Management Information Retrieval and Classification Algorithm

Efficient human resource management (HRM) is essential for company achievement in today’s fast-paced corporate world. Businesses must find effective ways to retrieve and categorise the ever-growing amount of HR-related knowledge. This work presents a new method for retrieving and classifying HRM data using machine learning. Modern natural language processing (NLP) and CNN methods are used by the algorithm we developed to handle unorganized human resources (HR) information, including worker records, job postings, and certificates. HR decision-making is facilitated by the system’s ability to derive insightful information from the information using sophisticated text mining and machine learning algorithms.The two main parts of the method are classification and data extraction. HR workers can more easily obtain the required knowledge thanks to information retrieval, which makes it possible to search HR data quickly and accurately. Contrarily, categorisation optimises the division of human resources information into pertinent classifications, including job positions, competencies, and achievement grades. We assess our algorithm’s effectiveness on various datasets from actual HR datasets. The outcomes show how well our strategy works to streamline HRM procedures, cut down on manual labour, and increase the precision of decisions. Furthermore, our technology is compatible with corporate human resources offices and educational settings because it complies with worldwide university requirements.This study belongs to the growing body of knowledge in HRM. It provides a useful tool for businesses looking to improve employee relations, simplify HR procedures, and attract and retain talent. The suggested method is a valuable resource for academics and businesses alike because of its versatility and compliance with global educational norms.

Leveraging natural language processing models to automate speech-intelligibility scoring

ABSTRACT Assessment of speech intelligibility in noise is critical for measuring the impact of age-related hearing loss. However, quantifying intelligibility often requires a human to manually process responses provided by a participant or patient to obtain a speech-intelligibility score – typically the proportion of correctly heard words. This manual process can be time-consuming and thus costly. The current study investigates whether state-of-the-art Natural Language Processing (NLP) models from Google and OpenAI could be used to calculate speech-intelligibility scores as an alternative to human scoring. It was specifically tested whether NLP models capture common speech-in-noise perception phenomena in younger and older adults (N = 144) listening to speech masked by modulated or unmodulated babble noise. The results show that NLP speech-intelligibility scores closely matched intelligibility scores from a human scorer (r ∼0.95). The main difference is, on average, ∼2% underestimation of NLP intelligibility scores relative to human intelligibility scores for moderate to high signal-to-noise ratios. This underestimation results from participants making minor errors related to misspellings, gender, or tense, to which NLP models are sensitive, but human scorers typically correct prior to scoring. Critically, NLP models capture the known age-related reduction in intelligibility and the age-related reduction in the benefit from a modulated relative to an unmodulated masker. OpenAI’s ADA2 appears to perform the best out of the tested NLP models, showing no difference in the speech-in-noise phenomena compared to human scoring. The current study suggests that modern NLP models can be used to score speech-intelligibility data.

Readability and Writing Quality in Radiation Oncology Journal Articles from 2004 to 2024

PurposeScientific literature is a vital tool which we rely upon to communicate the findings of our studies; however, we rarely direct our study to the writing itself. Methods and MaterialsHere, we make use of modern natural language processing algorithms coupled with the large, open access PubMed Central corpus to analyze trends in writing complexity within the field of radiation oncology from 2004–2024. Changes in 1) required grade-level to comprehend, 2) lexical complexity, and 3) information content were assessed. Articles were also classified—and then analyzed—by disease subsite. ResultsWe found significant increases in the three domains over the twenty-year collection period. Genitourinary literature had the greatest readability while gastrointestinal literature was the most complex. ConclusionsThis analysis reveals broad increases in the complexity of our writing. This study demonstrates metrics to use, and benchmark values to refer to, to evaluate the complexity of radiation oncology journal articles.

Алгоритмы поиска вербальных маркеров идентичности в современном научном дискурсе

The article is devoted to the study of identity verbalization specifics via Data Mining. The research material consists of English texts from Internet scientific repositories and e-libraries devoted to various concepts of youth identity. A methodology based on the use of modern natural language processing and machine learning tools was developed and tested as part of the research. The analysis was carried out using the Natural Language Toolkit library for tokenization and POS-tagging procedures for calculating the frequency of tokens from the «identity» environment. Word Embeddings, pre-trained Word2Vec model and K-means algorithm were used for the subsequent analysis and clustering of words based on their semantic proximity. Gensim library and Scikit-learn library were used to work with the Word2Vec model. As a result, it was proved that in English scientific discourse young person’s identity is verbalized within 9 semantic categories: behavior, communities, communication, education, identity, language, practice, complexity, science, the most common of which are education (33%), language (21%) and communities (18%). N-grams analysis made it possible to identify semantic fields, establish their attributes, and evaluate texts’ similarity, which provided the most accurate vector space search for semantically close n-grams. Optimization made it possible to establish a similarity measure to rank phrases according to the query, as well as assign each n-gram a certain ranking weight. Improvements can be achieved by adding statistical word weighting, such as TF-IDF. The proposed system is capable of searching in a large text array of related phrases with a similar meaning.

Using of NLP Methods in Intelligent Educational Systems

For the effective organisation of educational processes supported by relevant intelligent learning systems, it is important to choose the right technologies that would ensure individualisation of learning, adequate perception of learning content, and the so-called “understanding” of texts in Ukrainian provided by students (description of the solution to a task, answers provided in their own words, not selected from the test answer options, questions to the system, etc.), prototyping, constant iteration during natural language text recognition and processing, and maximum reliability and efficiency of learning processes. The purpose of the article is to study and analyse various methods of natural language processing, and the concept of NLP, and to consider common problems and prospects for developing a software product for processing Ukrainian-language text in online courses that support intelligent learning systems based on it. The research methods are the main methodological approaches and technological tools for analysing natural language texts in intelligent educational systems and developing a system for supporting NLP (Natural Language Processing) technology in the linguistic analysis of texts in Ukrainian. Such methods include, in particular: systemic and comparative analyses to identify the features of intelligence and information (with elements of intellectualisation) systems; the method of expert evaluation, which involves the study of literary sources and information resources, interviews and surveys of experts, as well as the processes of developing and testing intelligent and information systems. The novelty of the study is the analysis of modern technologies for the development of online educational process support systems through the organisation of processes of perception of information provided by students in natural language, the results of which can be used in the development of their software product to support the educational process in Ukrainian, ensuring the improvement of learning efficiency through the use of NLP technology in the process of studying the relevant academic content. Conclusions. The paper analyses modern NLP methods. The analysis has led to the selection of tokenisation, normalisation, stemming and lemmatisation methods for use in intelligent learning systems in the linguistic analysis of the so-called “free” communication in the natural (Ukrainian) language of students in the process of studying the educational content of online courses. During the tokenisation of Ukrainian-language texts, we solved such problems as eliminating so-called “merged” tokens, correcting spelling mistakes, identifying common prefixes in compound words and their impact on the semantics of the corresponding lexemes, identifying common prefixes in abbreviations, and bringing words to their normal form. Lemmatisation is especially important for the Ukrainian language (with its large number of cases of nouns, adjectives, word forms, etc.) and it requires the use of specially compiled dictionaries of the subject area under consideration. In these dictionaries, word forms are presented in the forms of lemmas (i.e., nouns are presented in the nominative case).

Part-of-speech tagger for Bodo language using deep learning approach

Abstract Language processing systems such as part-of-speech (POS) tagging, named entity recognition, machine translation, speech recognition, and language modeling have been well-studied in high-resource languages. Nevertheless, research on these systems for several low-resource languages, including Bodo, Mizo, Nagamese, and others, is either yet to commence or is in its nascent stages. Language model (LM) plays a vital role in the downstream tasks of modern natural language processing. Extensive studies are carried out on LMs for high-resource languages. However, these low-resource languages are still underreprese. In this study, we first present BodoBERT, an LM for the Bodo language. To the best of our knowledge, this work is the first such effort to develop an LM for Bodo. Second, we present an ensemble deep learning-based POS tagging model for Bodo. The POS tagging model is based on combinations of BiLSTM with conditional random field and stacked embedding of BodoBERT with BytePairEmbeddings. We cover several LMs in the experiment to see how well they work in POS tagging tasks. The best-performing model achieves an F1 score of 0.8041. A comparative experiment was also conducted on Assamese POS taggers, considering that the language is spoken in the same region as Bodo.

Exploratory analysis on the natural language processing models for task specific purposes

Natural language processing (NLP) is a technology that has become widespread in the area of human language understanding and analysis. A range of text processing tasks such as summarisation, semantic analysis, classification, question-answering, and natural language inference are commonly performed using it. The dilemma of picking a model to help us in our task is still there. It’s becoming an impediment. This is where we are trying to determine which modern NLP models are better suited for the tasks set out above in order to compare them with datasets like SQuAD and GLUE. For comparison, BERT, RoBERTa, distilBERT, BART, ALBERT, and text-to-text transfer transformer (T5) models have been used in this study. The aim is to understand the underlying architecture, its effects on the use case and also to understand where it falls short. Thus, we were able to observe that RoBERTa was more effective against the models ALBERT, distilBERT, and BERT in terms of tasks related to semantic analysis, natural language inference, and question-answering. The reason is due to the dynamic masking present in RoBERTa. For summarisation, even though BART and T5 models have very similar architecture the BART model has performed slightly better than the T5 model.

Automated linguistic analysis in speech samples of Turkish-speaking patients with schizophrenia-spectrum disorders

Modern natural language processing (NLP) methods provide ways to objectively quantify language disturbances for potential use in diagnostic classification. We performed computerized language analysis in speech samples of 82 Turkish-speaking subjects, including 44 patients with schizophrenia spectrum disorders (SSD) and 38 healthy controls (HC). Exploratory analysis of speech samples involved 16 sentence-level semantic similarity features using SBERT (Sentence Bidirectional Encoder Representation from Text) as well as 8 generic and 8 part-of-speech (POS) features. The random forest classifier using SBERT-derived semantic similarity features achieved a mean accuracy of 85.6 % for the classification of SSD and HC. When semantic similarity features were combined with generic and POS features, the classifier's mean accuracy reached to 86.8 %. Our analysis reflected increased sentence-level semantic similarity scores in SSD. Generic and POS analyses revealed an increase in the use of verbs, proper nouns and pronouns in SSD while our results showed a decrease in the utilization of conjunctions, determiners, and both average and maximum sentence length in SSD compared to HC. Quantitative language features were correlated with the expressive deficit domain of BNSS (Brief Negative Symptom Scale) as well as with the duration of illness. These findings from Turkish-speaking interviews contribute to the growing evidence-based NLP-derived assessments in non-English-speaking patients.

Signs of imminent collapse: Can hotel bankruptcy or failure be predicted from guest reviews?

This study examines the conceptual and methodological potential of predicting imminent hotel failure based on detecting early warning linguistic signals encoded and embedded in user-generated hotel guest reviews. Facilitated by modern big data mining and natural language processing (NLP) methods the study extracts and analyzes topics, sentiments, and linguistic features by comparing reviews of failed hotels versus a comparison control group. The study implemented machine learning models to detect early warning signals presaging the impending closure, bankruptcy, or failure of hotels. Guided by principles underlying linguistic signaling (Spence, 1978) and signal detection theories (Pastore & Scheirer, 1974) as well as incorporating temporal dimension (time-to-failure) in the study’s methods, findings suggest that certain linguistic cues and features (topics, sentiments, and specific “code” words) from guest reviews at certain prior periods can reliably predict looming hotel failure, giving hotel managers the opportunity to avert it.

Automating the assessment of multicultural orientation through machine learning and natural language processing.

Recent scholarship has highlighted the value of therapists adopting a multicultural orientation (MCO) within psychotherapy. A newly developed performance-based measure of MCO capacities exists (MCO-performance task [MCO-PT]) in which therapists respond to video-based vignettes of clients sharing culturally relevant information in therapy. The MCO-PT provides scores related to the three aspects of MCO: cultural humility (i.e., adoption of a nonsuperior and other-oriented stance toward clients), cultural opportunities (i.e., seizing or making moments in session to ask about clients' cultural identities), and cultural comfort (i.e., therapists' comfort in cultural conversations). Although a promising measure, the MCO-PT relies on labor-intensive human coding. The present study evaluated the ability to automate the scoring of the MCO-PT transcripts using modern machine learning and natural language processing methods. We included a sample of 100 participants (n = 613 MCO-PT responses). Results indicated that machine learning models were able to achieve near-human reliability on the average across all domains (Spearman's ρ = .75, p < .0001) and opportunity (ρ = .81, p < .0001). Performance was less robust for cultural humility (ρ = .46, p < .001) and was poorest for cultural comfort (ρ = .41, p < .001). This suggests that we may be on the cusp of being able to develop machine learning-based training paradigms that could allow therapists opportunities for feedback and deliberate practice of some key therapist behaviors, including aspects of MCO. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Harnessing explainable artificial intelligence for patient-to-clinical-trial matching: A proof-of-concept pilot study using phase I oncology trials.

This study aims to develop explainable AI methods for matching patients with phase 1 oncology clinical trials using Natural Language Processing (NLP) techniques to address challenges in patient recruitment for improved efficiency in drug development. A prototype system based on modern NLP techniques has been developed to match patient records with phase 1 oncology clinical trial protocols. Four criteria are considered for the matching: cancer type, performance status, genetic mutation, and measurable disease. The system outputs a summary matching score along with explanations of the evidence. The outputs of the AI system were evaluated against the ground truth matching results provided by the domain expert on a dataset of twelve synthesized dummy patient records and six clinical trial protocols. The system achieved a precision of 73.68%, sensitivity/recall of 56%, accuracy of 77.78%, and specificity of 89.36%. Further investigation into the misclassified cases indicated that ambiguity of abbreviation and misunderstanding of context are significant contributors to errors. The system found evidence of no matching for all false positive cases. To the best of our knowledge, no system in the public domain currently deploys an explainable AI-based approach to identify optimal patients for phase 1 oncology trials. This initial attempt to develop an AI system for patients and clinical trial matching in the context of phase 1 oncology trials showed promising results that are set to increase efficiency without sacrificing quality in patient-trial matching.

Medical Information Extraction With NLP-Powered QABots: A Real-World Scenario.

The advent of computerized medical recording systems in healthcare facilities has made data retrieval tasks easier, compared to manual recording. Nevertheless, the potential of the information contained within medical records remains largely untapped, mostly due to the time and effort required to extract data from unstructured documents. Natural Language Processing (NLP) represents a promising solution to this challenge, as it enables the use of automated text-mining tools for clinical practitioners. In this work, we present the architecture of the Virtual Dementia Institute (IVD), a consortium of sixteen Italian hospitals, using the NLP Extraction and Management Tool (NEMT), a (semi-) automated end-to-end pipeline that extracts relevant information from clinical documents and stores it in a centralized REDCap database. After defining a common Case Report Form (CRF) across the IVD hospitals, we implemented NEMT, the core of which is a Question Answering Bot (QABot) based on a modern NLP model. This QABot is fine-tuned on thousands of examples from IVD centers. Detailed descriptions of the process to define a common minimum dataset, Inter-Annotator Agreement calculated on clinical documents, and NEMT results are provided. The best QABot performance show an Exact Match score (EM) of 78.1%, a F1-score of 84.7%, a Lenient Accuracy (LAcc) of 0.834, and a Mean Reciprocal Rank (MRR) of 0.810. EM and F1 scores outperform the same metrics obtained with ChatGPTv3.5 (68.9% and 52.5%, respectively). With NEMT the IVD has been able to populate a database that will contain data from thousands of Italian patients, all screened with the same procedure. NEMT represents an efficient tool that paves the way for medical information extraction and exploitation for new research studies.

LANGUAGE IDENTIFICATION FOR MULTILINGUAL MACHINE

Language Identification for Multilingual Machine Translation is a crucial component in modern natural language processing systems, enabling accurate and efficient translation across multiple languages. This paper presents a comprehensive approach to language identification that enhances the performance of multilingual machine translation systems.The proposed method utilizes advanced machine learning techniques to automatically detect the language of a given text with high accuracy. By incorporating a variety of linguistic features and leveraging large-scale multilingual datasets, the system can identify languages even in challenging scenarios such as code-switching and mixed-language inputs.Key contributions of this work include the development of a robust language identification model that integrates seamlessly with machine translation pipelines. The model is evaluated on diverse datasets, demonstrating its effectiveness in real-world applications. Additionally, we explore the impact of accurate language identification on the overall quality of machine translation, highlighting improvements in translation accuracy and fluency.

Performance Evaluation of Deep Learning Techniques in The Detection of IOT Malware

Internet of Things (IoT) equipment is rapidly being used in a variety of businesses and for a variety of reasons (for example, sensing and collecting data from the environment in both public and military settings). Because of their expanding involvement in a wide range of applications and their rising computational and processing capabilities, they are a viable attack target for malware tailored to infect specific IoT devices. This study investigates the potential of detecting IoT malware using different deep learning techniques: the classic feedforward neural network (FNN), convolutional neural networks (CNN), long short-term memory (LSTM), and recurrent neural networks (RNN). The proposed method analyses the execution operation codes of IOT app sequences using modern NLP (natural language processing) methods. The current work utilized an IoT application dataset with 500 malware (collected from the IOTPOT dataset) and 500 goodware samples to train the proposed algorithms. The trained model is tested against 2971 fresh IoT malware and goodware samples. The samples were input into deep learning models, and performance metrics were obtained. The results demonstrate that the RNN model had the best accuracy (99.19%) in detecting fresh malware samples. On the other hand, the results were compared by the time required for training; the CNN model shows that it could achieve high accuracy (98.05%) with less training time. A comparison with various deep learning classifiers demonstrates that the RNN and CNN techniques produce the best results.

NLP and its Components: A Detailed Discussion

Natural Language Processing (NLP) encompasses computational techniques for processing and analyzing human language, primarily through Natural Language Understanding (NLU) and Natural Language Generation (NLG). NLU focuses on interpreting language by analyzing phonology (sounds), morphology (word structures), syntax (sentence structures), semantics (meaning), and pragmatics (context). These processes enable machines to comprehend the nuances of human language for accurate interpretation and response generation. NLG, in contrast, involves producing human-like text from structured data. It includes content determination (identifying relevant information), text planning (organizing information), sentence planning (constructing grammatically correct sentences), and surface realization (generating the final text). NLG is crucial for applications like automated report generation and chat bots, where coherent and contextually appropriate responses are essential. The synergy between NLU and NLG underpins many NLP applications. In machine translation, NLU interprets the source text, while NLG generates the translated text. In question-answering systems, NLU processes the query, and NLG formulates the response. Deep learning advancements have significantly enhanced both NLU and NLG, enabling more sophisticated and human-like interactions. This paper explores the components and processes of NLU and NLG, offering a comprehensive understanding of their mechanisms and the advancements driving modern NLP systems.

Toward Machine-learning-based Metastudies: Applications to Cosmological Parameters

We develop a new model for automatic extraction of reported measurement values from the astrophysical literature, utilizing modern natural language processing techniques. We use this model to extract measurements present in the abstracts of the approximately 248,000 astrophysics articles from the arXiv repository, yielding a database containing over 231,000 astrophysical numerical measurements. Furthermore, we present an online interface (Numerical Atlas) to allow users to query and explore this database, based on parameter names and symbolic representations, and download the resulting data sets for their own research uses. To illustrate potential use cases, we then collect values for nine different cosmological parameters using this tool. From these results, we can clearly observe the historical trends in the reported values of these quantities over the past two decades and see the impacts of landmark publications on our understanding of cosmology.

Development of an intelligent information resource model based on modern natural language processing methods

&lt;span lang="EN-US"&gt;Currently, there is an avalanche-like increase in the need for automatic text processing, respectively, new effective methods and tools for processing texts in natural language are emerging. Although these methods, tools and resources are mostly presented on the internet, many of them remain inaccessible to developers, since they are not systematized, distributed in various directories or on separate sites of both humanitarian and technical orientation. All this greatly complicates their search and practical use in conducting research in computational linguistics and developing applied systems for natural text processing. This paper is aimed at solving the need described above. The paper goal is to develop model of an intelligent information resource based on modern methods of natural language processing (IIR NLP). The main goal of IIR NLP is to render convenient valuable access for specialists in the field of computational linguistics. The originality of our proposed approach is that the developed ontology of the subject area “NLP” will be used to systematize all the above knowledge, data, information resources and organize meaningful access to them, and semantic web standards and technology tools will be used as a software basis.&lt;/span&gt;

Feasibility of Using Zero-Shot Learning in Transformer-Based Natural Language Processing Algorithm for Key Information Extraction from Head and Neck Tumor Board Notes

Natural language processing (NLP) technology has the potential to automate information aggregation and summarization in oncology. One example is the automation of patient registry creation. In this work, we aim to show (1) the feasibility of using modern NLP algorithms to extract key information from tumor board notes, and (2) the impact of prompt engineering on the quality of the results. In this IRB-approved study, we obtained the texts of head and neck tumor board notes for 306 unique patients. Five key pieces of information used to create a patient registry were predefined: age, gender, tumor histology, tumor stage, and primary location. The NLP algorithm used was a modified Text-To-Text Transfer Transformer (T5) model that was initially trained on the Colossal Clean Crawled Corpus (C4) dataset and subsequently fine-tuned on the Stanford Question Answering Dataset (SQuAD) to perform the downstream task of extractive question answering. The NLP model and trained weights were obtained from the Hugging Face platform. During inference, the entire body of the tumor board note and a related question were fed as inputs, and the model predicted a sequence of texts in response to the question. Two sets of questions of similar semantic meanings were used. Questions in prompt set #1 included "What is the gender?", "What is the age?", "What is the type of carcinoma in pathological diagnosis?", "What is the stage?", and "Where is the carcinoma located at?". Questions in prompt set #2 include "Is the patient male or female?", "How old is the patient?", "What kind of cancer?", "What is the cancer stage?", and "What is the tumor location?". Each model-predicted response was compared to the ground truth extracted from the tumor board notes. A response was classified as true if it is consistent with the ground truth, otherwise, it was deemed false. The response accuracy for each question was subsequently calculated. The median number of words in each tumor board note was 448 (range, 219 - 1505). The accuracy of the NLP algorithm for each question from either set is reported in Table 1. Algorithm performance is higher for extracting objective information such as age, gender, and histology. In addition, it was found that questions of similar semantic meanings but with different wording can lead to significantly different results. We demonstrated that a transformer-based extractive question-answering NLP algorithm can be successfully used for extracting information from head and neck tumor board notes with zero-shot learning. Furthermore, our results highlight the significance of prompt engineering for applying NLP for this task. Future work on finetuning these algorithms to oncology-specific texts can potentially enhance algorithm performance for more difficult tasks.