Search results

User feedback inferred from the user's search-time behavior could improve the learning to rank (L2R) algorithms. Click models (CMs) present probabilistic frameworks for describing and predicting the user's clicks during search sessions. Most of these CMs are based on common assumptions such as Attractiveness, Examination and User Satisfaction. CMs usually consider the Attractiveness and Examination as pre- and post-estimators of the actual relevance. They also assume that User Satisfaction is a function of the actual relevance. This paper extends the authors' previous work by building a reinforcement learning (RL) model to predict the relevance. The Attractiveness, Examination and User Satisfaction are estimated using a limited number of the features of the utilized benchmark data set and then they are incorporated in the construction of an RL agent. The proposed RL model learns to predict the relevance label of documents with respect to a given query more effectively than the baseline RL models for those data sets.

In this paper, User Satisfaction is used as an indication of the relevance level of a query to a document. User Satisfaction itself is estimated through Attractiveness and Examination, and in turn, Attractiveness and Examination are calculated by the random forest algorithm. In this process, only a small subset of top information retrieval (IR) features are used, which are selected based on their mean average precision and normalized discounted cumulative gain values. Based on the authors' observations, the multiplication of the Attractiveness and Examination values of a given query–document pair closely approximates the User Satisfaction and hence the relevance level. Besides, an RL model is designed in such a way that the current state of the RL agent is determined by discretization of the estimated Attractiveness and Examination values. In this way, each query–document pair would be mapped into a specific state based on its Attractiveness and Examination values. Then, based on the reward function, the RL agent would try to choose an action (relevance label) which maximizes the received reward in its current state. Using temporal difference (TD) learning algorithms, such as Q-learning and SARSA, the learning agent gradually learns to identify an appropriate relevance label in each state. The reward that is used in the RL agent is proportional to the difference between the User Satisfaction and the selected action.

Experimental results on MSLR-WEB10K and WCL2R benchmark data sets demonstrate that the proposed algorithm, named as SeaRank, outperforms baseline algorithms. Improvement is more noticeable in top-ranked results, which usually receive more attention from users.

This research provides a mapping from IR features to the CM features and thereafter utilizes these newly generated features to build an RL model. This RL model is proposed with the definition of the states, actions and reward function. By applying TD learning algorithms, such as the Q-learning and SARSA, within several learning episodes, the RL agent would be able to learn how to choose the most appropriate relevance label for a given pair of query–document.

Prediction models are essential tools for transportation agencies to forecast the condition of bridge decks based on available data, and artificial intelligence is paramount for this purpose. This study aims at proposing a bridge deck condition prediction model by assessing various classification and regression algorithms.

The 2019 National Bridge Inventory database is considered for model development. Eight different feature selection techniques, along with their mean and frequency, are used to identify the critical features influencing deck condition ratings. Thereafter, four regression and four classification algorithms are applied to predict condition ratings based on the selected features, and their performances are evaluated and compared with respect to the mean absolute error (MAE).

Classification algorithms outperform regression algorithms in predicting deck condition ratings. Due to its minimal MAE (0.369), the random forest classifier with eleven features is recommended as the preferred condition prediction model. The identified dominant features are superstructure condition, age, structural evaluation, substructure condition, inventory rating, maximum span length, deck area, average daily traffic, operating rating, deck width, and the number of spans.

The proposed bridge deck condition prediction model offers a valuable tool for transportation agencies to plan maintenance and resource allocation efficiently, ultimately improving bridge safety and serviceability.

This study provides a detailed framework for applying machine learning in bridge condition prediction that applies to any bridge inventory database. Moreover, it uses a comprehensive dataset encompassing an entire region, broadening the model’s applicability and representation.

For ranking aggregation in crowdsourcing task, the key issue is how to select the optimal working group with a given number of workers to optimize the performance of their aggregation. Performance prediction for ranking aggregation can solve this issue effectively. However, the performance prediction effect for ranking aggregation varies greatly due to the different influencing factors selected. Although questions on why and how data fusion methods perform well have been thoroughly discussed in the past, there is a lack of insight about how to select influencing factors to predict the performance and how much can be improved of.

In this paper, performance prediction of multivariable linear regression based on the optimal influencing factors for ranking aggregation in crowdsourcing task is studied. An influencing factor optimization selection method based on stepwise regression (IFOS-SR) is proposed to screen the optimal influencing factors. A working group selection model based on the optimal influencing factors is built to select the optimal working group with a given number of workers.

The proposed approach can identify the optimal influencing factors of ranking aggregation, predict the aggregation performance more accurately than the state-of-the-art methods and select the optimal working group with a given number of workers.

To find out under which condition data fusion method may lead to performance improvement for ranking aggregation in crowdsourcing task, the optimal influencing factors are identified by the IFOS-SR method. This paper presents an analysis of the behavior of the linear combination method and the CombSUM method based on the optimal influencing factors, and optimizes the task assignment with a given number of workers by the optimal working group selection method.

This paper aims to manage the word gap in information retrieval (IR) especially for long documents belonging to specific domains. In fact, with the continuous growth of text data that modern IR systems have to manage, existing solutions are needed to efficiently find the best set of documents for a given request. The words used to describe a query can differ from those used in related documents. Despite meaning closeness, nonoverlapping words are challenging for IR systems. This word gap becomes significant for long documents from specific domains.

To generate new words for a document, a deep learning (DL) masked language model is used to infer related words. Used DL models are pretrained on massive text data and carry common or specific domain knowledge to propose a better document representation.

The authors evaluate the approach of this study on specific IR domains with long documents to show the genericity of the proposed model and achieve encouraging results.

In this paper, to the best of the authors’ knowledge, an original unsupervised and modular IR system based on recent DL methods is introduced.

The purpose of this paper is to explore the factors influencing the encountering information adoption of virtual live streaming from the perspective of the immersion experience. In addition, the paper aims to provide new theoretical perspectives and analytical frameworks for virtual live information behavior.

Based on a review of relevant literature and theories, a model of the encountering information adoption of virtual live streaming users is constructed. In order to complete the empirical study, two experiments and questionnaires have been designed to investigate the relationship between high and low immersion experiences. A total of 1,332 valid survey samples were collected and analyzed, utilizing the structural equation model. In order to delineate the regimes, Gradient Boosted Regression Tree (GBRT) and Lasso regression were further utilized.

The research findings indicate that users' immersion experience in virtual live streaming has a positive effect on perceived usefulness, trust, and commitment. Furthermore, perceived usefulness and trust have a positive effect on users' emotional arousal and enhance the content experience, while commitment has a negative effect on the content experience. The emotional arousal and content experience of users contribute to their encountering information adoption. The effect of immersion experience on encountering information adoption is partially mediated by perceived usefulness, trust, commitment, emotional arousal, and content experience. The relationship between content experience and encountering information adoption is moderated by digital literacy to a significant extent. In the context of virtual live streaming, the factors influencing users' encountering information adoption can be divided into three distinct regimes. The most significant factors affecting encounter information adoption are trust and commitment, which are located in the first regime. Emotional arousal and digital literacy are situated in the third regime, with the least significant influence on encountering information adoption.

This study constructs a model of virtual live streaming users' encountering information adoption and explores the formation mechanism of encountering information adoption from the perspective of immersion experience, which provides a new perspective for further understanding the influence of virtual live-streaming users' encountering information adoption.

With the development of intelligent technology, deep learning has made significant progress and has been widely used in various fields. Deep learning is data-driven, and its training process requires a large amount of data to improve model performance. However, labeled data is expensive and not readily available.

To address the above problem, researchers have integrated semi-supervised and deep learning, using a limited number of labeled data and many unlabeled data to train models. In this paper, Generative Adversarial Networks (GANs) are analyzed as an entry point. Firstly, we discuss the current research on GANs in image super-resolution applications, including supervised, unsupervised, and semi-supervised learning approaches. Secondly, based on semi-supervised learning, different optimization methods are introduced as an example of image classification. Eventually, experimental comparisons and analyses of existing semi-supervised optimization methods based on GANs will be performed.

Following the analysis of the selected studies, we summarize the problems that existed during the research process and propose future research directions.

This paper reviews and analyzes research on generative adversarial networks for image super-resolution and classification from various learning approaches. The comparative analysis of experimental results on current semi-supervised GAN optimizations is performed to provide a reference for further research.

This study aimed to investigate how the enrichment of medical documents' index terms by their comments improves the relevance and novelty of the top-ranked results retrieved by an NLP system.

A semi-experimental pre-test and post-test research was designed to compare NLP-based indexes before and after being expanded by the comment terms. The experiments were conducted on a test collection of 13,957 documents commented by F1000-Prime reviewers. They were indexed at title, abstract, body and full-text levels. In total, 100 seed documents were randomly selected and served as queries. The textual similarity of the documents and queries was calculated using Lucene-more-like-this function and evaluated by the semantic similarity of their MeSH. The results novelty was measured using maximal marginal relevance and evaluated by their MeSH novelties. Normalized discounted cumulative gain was used to compare the basic and expanded indexes' precisions at 10, 20 and 50 top ranks.

The relevance and novelty of the results ranked at the top precision points was improved after expanding the indexes by the comment terms. The finding implies that meta-texts are effective in representing their mother documents, by adding dynamic elements to their rather static contents. It also provides further evidence about the merits of the application of social intelligence and collective wisdom reflected in the actions and reactions of users in tackling the challenges faced by NLP-based systems.

This is the first study to confirm that social comments on scientific papers improve the performance of information systems in terms of relevance and novelty.

The peer review history for this article is available at: https://publons.com/publon/10.1108/OIR-05-2022-0283.

The purpose of this study is to examine the impact of country-level governance on sustainability performance, taking into account the effect of sustainable development goals (SDGs) and board characteristics.

This study uses panel data analysis using fixed effect models to investigate the influence of country-level governance on sustainability performance while considering the effect of SDGs and board characteristics. The sample comprises 8,273 firms across 41 countries during the period spanning from 2016 to 2021. The sample is divided into two categories based on the score of SDGs.

The findings of this study show that countries with high SDGs score have better overall country-level governance and board attributes which have a statistically significant positive impact on sustainability performance. However, for those countries with low SDGs, political stability shows a statistically insignificant and negative impact on sustainability performance, while government effectiveness indicates a statistically insignificant positive impact on sustainability performance.

This study contributes to the literature by providing empirical evidence on the relationship between country-level governance, SDGs, board characteristics and sustainability performance. The study also highlights the importance of considering the effect of SDGs on the relationship between country-level governance and sustainability performance. The findings of this study could be useful for policymakers and firms in improving their sustainability performance and contributing to sustainable development.

Histopathology biopsy imaging is currently the gold standard for the diagnosis of breast cancer in clinical practice. Pathologists examine the images at various magnifications to identify the type of tumor because if only one magnification is taken into account, the decision may not be accurate. This study explores the performance of transfer learning and late fusion to construct multi-scale ensembles that fuse different magnification-specific deep learning models for the binary classification of breast tumor slides.

Three pretrained deep learning techniques (DenseNet 201, MobileNet v2 and Inception v3) were used to classify breast tumor images over the four magnification factors of the Breast Cancer Histopathological Image Classification dataset (40×, 100×, 200× and 400×). To fuse the predictions of the models trained on different magnification factors, different aggregators were used, including weighted voting and seven meta-classifiers trained on slide predictions using class labels and the probabilities assigned to each class. The best cluster of the outperforming models was chosen using the Scott–Knott statistical test, and the top models were ranked using the Borda count voting system.

This study recommends the use of transfer learning and late fusion for histopathological breast cancer image classification by constructing multi-magnification ensembles because they perform better than models trained on each magnification separately.

The best multi-scale ensembles outperformed state-of-the-art integrated models and achieved an accuracy mean value of 98.82 per cent, precision of 98.46 per cent, recall of 100 per cent and F1-score of 99.20 per cent.

This study examined the factors impeding the implementation of micro-credentials and accepting it as a credible source of earning professional qualifications and certifications necessary for pursuing higher education or other career goals.

The factors were identified by reflecting on the recent literature and Internet resources coupled with in-depth brainstorming with experts in the field of micro-credentials including educators, learners and employers. Two ranking methods, namely Preference Ranking for Organization Method for Enrichment Evaluation (PROMETHEE) and multi-objective optimization based on ratio analysis (MOORA), are used together to rank the major challenges.

The results of this study present that lack of clear definitions, ambiguous course descriptions, lack of accreditation and quality assurance, unclear remuneration policies, lack of coordination between learning hours and learning outcomes, the inadequate volume of learning, and lack of acceptance by individuals and organizations are the top-ranked and the most significant barriers in the implementation of micro-credentials.

The findings can be used by educational institutions, organizations and policymakers to better understand the issues and develop strategies to address them, making micro-credentials a more recognized form of education and qualifications.

The novelty of this study is to identify the primary factors influencing the implementation of micro-credentials from the educators', students' and employers' perspectives and to prioritize those using ranking methods such as PROMETHEE and MOORA.