Exploring topic continuation follow-up questions using machine learning

Question-answering systems are recognized as popular and frequently effective means of information seeking on the web. In such systems, information seekers can receive a concise response to their queries by presenting their questions in natural language. Interactive question answering is a recently proposed and increasingly popular solution that resides at the intersection of question answering and dialogue systems . On the one hand, the user can ask questions in normal language and locate the actual response to her inquiry; on the other hand, the system can prolong the question-answering session into a dialogue if there are multiple probable replies, very few, or ambiguities in the initial request. By permitting the user to ask more questions, interactive question answering enables users to interact with the system and receive more precise results dynamically. This survey offers a detailed overview of the interactive question-answering methods that are prevalent in current literature. It begins by explaining the foundational principles of question-answering systems, hence defining new notations and taxonomies to combine all identified works inside a unified framework. The reviewed published work on interactive question-answering systems is then presented and examined in terms of its proposed methodology, evaluation approaches, and dataset/application domain. We also describe trends surrounding specific tasks and issues raised by the community, so shedding light on the future interests of scholars. Our work is further supported by a GitHub page synthesizing all the major topics covered in this literature study. https://sisinflab.github.io/interactive-question-answering-systems-survey/

Evaluating interactive question answering (QA) systems with real users can be challenging because traditional evaluation measures based on the relevance of items returned are difficult to employ since relevance judgments can be unstable in multi-user evaluations. The work reported in this paper evaluates, in distinguishing among a set of interactive QA systems, the effectiveness of three questionnaires: a Cognitive Workload Questionnaire (NASA TLX), and Task and System Questionnaires customized to a specific interactive QA application. These Questionnaires were evaluated with four systems, seven analysts, and eight scenarios during a 2-week workshop. Overall, results demonstrate that all three Questionnaires are effective at distinguishing among systems, with the Task Questionnaire being the most sensitive. Results also provide initial support for the validity and reliability of the Questionnaires.

Improving autocoding performance of rare categories in injury classification: Is more training data or filtering the solution?

Search engines can return ranked documents as a result for any query from which the user struggle to navigate and search the correct answer. This process wastes user's navigation time and due to this the need for automated question answering systems becomes more urgent. We need such a system which is capable of replying the exact and concise answer to the question posed in natural language. The best way to address this problem is use of Question answering systems (QAS). The basic aim of QAS is to provide short and correct answer to the user saving his/her navigation time. The concept of Natural Language Processing plays an important role in developing any QAS. This paper provides an implementation approaches for various categories of QAS such as Closed Domain based QAS, Open Domain based QAS, WEBBASED QAS, Information Retrieval or Information Extraction (IR/IE) based QAS, and Rule based QAS which will be helpful for new directions of research in this area.

Comparative study of different machine learning models in landslide susceptibility assessment: A case study of Conghua District, Guangzhou, China

Assessing treatment switch among patients with multiple sclerosis: A machine learning approach

Background/Objective: Subarachnoid hemorrhage (SAH) is associated with high morbidity and mortality rates, necessitating prognostic algorithms to guide decisions. Our study evaluates the use of machine learning (ML) models for predicting 1-month and 1-year mortality among SAH patients using national electronic health records (EHR) system. Methods: Retrospective cohort of 29,274 SAH patients, identified through national EHR system from January 2017 to December 2022, was analyzed, with mortality data obtained from central civil registration system in Türkiye. Variables included (n = 102) pre- (n = 65) and post-admission (n = 37) data, such as patient demographics, clinical presentation, comorbidities, laboratory results, and complications. We employed logistic regression (LR), decision trees (DTs), random forests (RFs), and artificial neural networks (ANN). Model performance was evaluated using area under the curve (AUC), average precision, and accuracy. Feature significance analysis was conducted using LR. Results: The average age was 56.23 ± 16.45 years (47.8% female). The overall mortality rate was 22.8% at 1 month and 33.3% at 1 year. One-month mortality increased from 20.9% to 24.57% (p < 0.001), and 1-year mortality rose from 30.85% to 35.55% (p < 0.001) in the post-COVID period compared to the pre-COVID period. For 1-month mortality prediction, the ANN, LR, RF, and DT models achieved AUCs of 0.946, 0.942, 0.931, and 0.916, with accuracies of 0.905, 0.901, 0.893, and 0.885, respectively. For 1-year mortality, the AUCs were 0.941, 0.927, 0.926, and 0.907, with accuracies of 0.884, 0.875, 0.861, and 0.851, respectively. Key predictors of mortality included age, cardiopulmonary arrest, abnormal laboratory results (such as abnormal glucose and lactate levels) at presentation, and pre-existing comorbidities. Incorporating post-admission features (n = 37) alongside pre-admission features (n = 65) improved model performance for both 1-month and 1-year mortality predictions, with average AUC improvements of 0.093 ± 0.011 and 0.089 ± 0.012, respectively. Conclusions: Our study demonstrates the effectiveness of ML models in predicting mortality in SAH patients using big data. LR models’ robustness, interpretability, and feature significance analysis validate its importance. Including post-admission data significantly improved all models’ performances. Our results demonstrate the utility of big data analytics in population-level health outcomes studies.

Machine Learning Can Accurately Predict Overnight Stay, Readmission, and 30-Day Complications Following Anterior Cruciate Ligament Reconstruction

This paper presents a Retrieval-Augmented Generation (RAG) system for question answering over the set of documents of a public tender. The proof-of-concept chatbot is intended to provide support to officers participating in the evaluation process of a tender whose documents are indexed. The system implements a dual-mode retrieval strategy —supporting both semantic and lexical search— to extract relevant context and uses an instruction-following language model to generate grounded responses based strictly on the retrieved information. The system is designed to assist officers in exploring tenders, evaluation reports, and award justifications through natural language queries. In order to assess the system, the documents of a tender have been generated symtheticly. Qualitative testing demonstrated that the system provided accurate, verifiable answers, avoiding unsupported inferences. The architecture emphasizes document fidelity, interpretability, and usability for non-technical users involved in procurement oversight. Future work may extend the framework to support the identification of fraud-related patterns.

The goal of the question-answering (QA) system is to understand the questions from users and return their accurate answers. In the medical field, the question-answering system amis to understand patients' questions and return the correct answers. The existed knowledge base question-answering (KB-QA) systems mainly rely on hand-crafted features and ignore structure information of knowledge base which accordingly lead to the answers with low accuracy. In this paper, a novel KB-QA model is put forward based on knowledge representation and recurrent convolutional neural network. This model has three parts, candidate answers generation, entity relationships extraction and knowledge representation learning based on knowledge base. In addition, an algorithm is also developed to compute the scores of linking candidate answers and knowledge base. Experimental results show that the presented model achieves better performance compared with the baseline systems.

This paper presents an extension of the community question-answering (CQA) system we developed previously. This graph-based system method that builds ranked answers for related questions using nKullback–Leibler (KL) divergence. The process of extracting answers to questions in this work involves; question core, building question query, query-based answer extraction (QBAE), pattern-based answer extraction (PBAE), and combined answer extraction. The source data for this work were existing data from ResearchGate, a socio-academic networking website that provides researchers the platform to collaborate, ask questions, and offer answers to questions. The performance for answer extraction for 2786 questions shows that when 80% of patterns and keywords were considered, QBAE and PBAE extracted 2765 and 2766 correct answers, respectively, while the QBAE + PBAE method extracted 2782 correct answers. Also, when 90% of patterns and keywords were utilized, QBAE and PBAE extracted 2782 and 2784 correct answers, whereas the QBAE + PBAE method extracted 2786 correct answers. Our method was able to identify 229 questions without answers. Finally, the evaluation of our model reveals high-performance accuracy and precision.

Watson, the winner of the Jeopardy! challenge, is a state-of-the-art open-domain Question Answering system that tackles the fundamental issue of answer typing by using a novel type coercion (TyCor) framework, where candidate answers are initially produced without considering type information, and subsequent stages check whether the candidate can be coerced into the expected answer type. In this paper, we provide a high-level overview of the TyCor framework and discuss how it is integrated in Watson, focusing on and evaluating three TyCor components that leverage the community built semi-structured and structured knowledge resources – DBpedia (in conjunction with the YAGO ontology), Wikipedia Categories and Lists. These resources complement each other well in terms of precision and granularity of type information, and through links to Wikipedia, provide coverage for a large set of instances.

The clinical necessity of lymph node dissection in papillary thyroid carcinoma (PTC) surgery remains contentious. This study compared four logistic regression (LR) models (with distinct feature selection strategies) and four machine learning (ML) models to preoperatively predict lymph node metastasis (LNM) risk in PTC patients, with emphasis on multidimensional evaluation and cross-populational generalizability. Data from 3,175 PTC patients (2021 cohort) were randomly split into training (70%) and testing (30%) subsets, with external validation performed using a Chinese (2024, n = 104) and a Canadian (2019–2022, n = 412) cohort. Twelve predictors were screened, and models were evaluated using metrics of discrimination (AUC), calibration (Brier Score), classification accuracy, and clinical utility. The prevalence of LNM was 34.48%, 36.54%, and 30.10% in the internal, Chinese, and Canadian cohorts, respectively. Among ML models, Random Forest achieved the highest internal AUC (0.767), whereas XGBoost demonstrated superior generalization (external AUCs: 0.785 and 0.725). LR models, particularly BestSubset-GLM, outperformed these ML models with an internal AUC of 0.770 and external AUCs of 0.831 and 0.785. Notably, BestSubset-GLM exhibited high specificity (0.86), precision (0.59), favorable calibration (Brier Score < 0.20), and robust clinical utility across the approximately 15–90% threshold probability range. Extrathyroidal extension, tumor size above 1.00 cm, younger age, and male gender were identified as key LNM risk factors. Bethesda classification and molecular aberrations were integrated into models. BestSubset-GLM balanced parsimony, interpretability, and generalizability, thereby supporting clinical decision-making through dynamic nomograms. Comprehensive evaluation beyond AUC is crucial.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-025-19345-4.

Short birth interval is an important reproductive health concern in low- and middle-income countries, including Bangladesh, as it may have adverse consequences for both perinatal and maternal health. According to the International Classification of Diseases, Tenth Revision (ICD-10) codes O09.891, O09.892 and O09.893, pregnancies occurring after a short interpregnancy interval are classified as high-risk pregnancies that require appropriate medical supervision. Understanding the risk factors associated with short birth intervals is crucial for designing effective interventions and achieving the maternal and child health targets outlined in the Sustainable Development Goals. Therefore, this study aimed to investigate the prevalence and identify the potential risk factors of short birth interval using a two-level logistic regression (LR) model and the Boruta machine learning (ML) based feature selection method. This was a cross-sectional study. The study used secondary data from the nationally representative Bangladesh Demographic and Health Survey (BDHS) 2022. The study included 11 872 married women of reproductive age (15-49 years) in Bangladesh who reported at least one previous live birth. The demographic, socioeconomic, reproductive and media exposure-related characteristics of the study participants were obtained from the BDHS 2022 dataset. A two-level LR model was used to identify the risk factors associated with short birth interval, and the results were presented as adjusted ORs (aORs) with 95% CIs and p values <0.05. Additionally, the Boruta ML method was applied to determine the important predictors of short birth intervals. The overlapping risk factors identified by both the LR model and the Boruta method were considered potential risk factors of short birth intervals among reproductive-aged women in Bangladesh. Subsequently, three widely used ML models were implemented in the study. The performance of each model was evaluated on the test dataset using five evaluation metrics, namely accuracy, precision, recall, F1-score and the area under the receiver operating characteristic curve. In Bangladesh, the prevalence of short birth intervals was 34%. Both the two-level LR model and the Boruta method identified women aged ≤19 years (aOR=6.496, 95% CI 2.900 to 14.551; p<0.001); residence in Sylhet (aOR=2.457, 95% CI 1.956 to 3.086; p<0.001), Chattogram (aOR=1.554, 95% CI 1.279 to 1.887; p<0.001), Dhaka (aOR=1.287, 95% CI 1.059 to 1.565; p=0.011), Mymensingh (aOR=1.394, 95% CI 1.118 to 1.738; p=0.003) and Rangpur (aOR=1.248, 95% CI 1.009 to 1.543; p=0.041) divisions; no education (aOR=1.399, 95% CI 1.078 to 1.816; p=0.011); poorest (aOR=1.364, 95% CI 1.146 to 1.624; p<0.001), poorer (aOR=1.230, 95% CI 1.042 to 1.452; p=0.014) and middle wealth quintile (aOR=1.196, 95% CI 1.022 to 1.399; p=0.025); use of traditional contraceptive methods (aOR=1.615, 95% CI 1.305 to 1.998; p<0.001) and no intention to use contraception (aOR=1.506, 95% CI 1.237 to 1.833; p<0.001); birth order ≥4 (aOR=1.467, 95% CI 1.243 to 1.730; p<0.001); and having more than two children ever born (aOR=2.777, 95% CI 2.455 to 3.141; p<0.001) as potential risk factors for short birth interval among reproductive-aged women in Bangladesh. Among the ML-based models, the extreme gradient boosting model achieved the highest performance, with an accuracy of 66.356% in predicting short birth intervals. The findings highlight that women aged ≤19 years, those residing in the Eastern and Central divisions, women with no formal education, lower household wealth quintile, traditional or no intention to use contraceptive methods, higher birth order and having more than two children were potential risk factors for short birth interval. The extreme gradient boosting classifier showed the best performance in predicting short birth intervals among reproductive-aged women in Bangladesh.

Prediction of non-suicidal self-injury in adolescents at the family level using regression methods and machine learning