Real-time Recommendation Research Articles

Recommendation of Scientific Publications—A Real-Time Text Analysis and Publication Recommendation System

Studies have shown that although having more information improves the quality of decision-making, information overload causes adverse effects on decision quality. Visual analytics and recommendation systems counter this adverse effect on decision-making. Accurately identifying relevant information can reduce the noise during exploration and improve decision-making. These countermeasures also help scientists make correct decisions during research. We present a novel and intuitive approach that supports real-time collaboration. In this paper, we instantiate our approach to scientific writing and propose a system that supports scientists. The proposed system analyzes text as it is being written and recommends similar publications based on the written text through similarity algorithms. By analyzing text as it is being written, it is possible to provide targeted real-time recommendations to improve decision-making during research by finding relevant publications that might not have been otherwise found in the initial research phase. This approach allows the recommendations to evolve throughout the writing process, as recommendations begin on a paragraph-based level and progress throughout the entire written text. This approach yields various possible use cases discussed in our work. Furthermore, the recommendations are presented in a visual analytics system to further improve scientists’ decision-making capabilities.

Real-Time Context-Aware Recommendation System for Tourism.

Recently, the tourism trend has been shifting towards the Tourism 2.0 paradigm due to increased travel experiences and the increase in acquiring and sharing information through the Internet. The Tourism 2.0 paradigm requires developing intelligent tourism service tools for positive effects such as time savings and marketing utilization. Existing tourism service tools recommend tourist destinations based on the relationship between tourists and tourist destinations or tourism patterns, so it is difficult to make recommendations in situations where information is insufficient or changes in real time. In this paper, we propose a real-time recommendation system for tourism (R2Tour) that responds to changing situations in real time, such as external factors and distance information, and recommends customized tourist destinations according to the type of tourist. R2Tour trains a machine learning model with situational information such as temperature and precipitation and tourist profiles such as gender and age to recommend the top five nearby tourist destinations. To verify the recommendation performance of R2Tour, six machine learning models, including K-NN and SVM, and information on tourist attractions in Jeju Island were used. As a result of the experiment, R2Tour was verified with accuracy of 77.3%, micro-F1 0.773, and macro-F1 0.415. Since R2Tour trains tourism patterns based on situational information, it is possible to recommend new tourist destinations and respond to changing situations in real time. In the future, R2Tour can be installed in vehicles to recommend nearby tourist destinations or expanded to tasks in the tourism industry, such as a smart target advertising system.

Real-time artificial intelligence assisted insulin dosage titration system for glucose control in type 2 diabetic patients: a proof of concept study

ObjectiveThis study aims to develop an insulin dosage adjustment model using machine learning of high quality electronic health records (EHRs) notes and then to form an artificial intelligence-based insulin clinical decision support workflow (iNCDSS) implemented in the HIS system to give a real-time recommendation of insulin dosage titration. The efficacy and safety in clinical practice is evaluated in this proof-of-concept study.Research design and methodsWe extracted patient-specific and time-varying features from the original EHRs data and performed machine learning analysis through 5-fold cross validation. In the patient-blind, single-arm interventional study, insulin dosage was titrated according to iNCDSS in type 2 diabetic inpatients for up to 7 d or until hospital discharge. The primary end point of the trial was the difference in glycemic control as measured by mean daily blood glucose concentration during the intervention period.ResultsA total of 3275 type 2 diabetic patients with 38,406 insulin counts were included for the model analysis. The XGBoost model presented the best performance with root mean square error (RMSE) of 1.06 unit and mean absolute relative difference (MARD) of 6.0% in the training dataset, and RMSE of 1.30 unit and MARD of 6.9% in the testing dataset. Twenty-three patients with T2DM (male 14, 60.9%; age 58.8 ± 10.7 years; duration of diabetes 11.8 ± 8.8 years, HbA1c 9.1 ± 1.1%) were enrolled in the proof of concept trial. The duration of iNCDSS intervention was 7.0 ± 0.1 d. The insulin dose recommended by iNCDSS was accepted by physicians in 97.8%. The mean daily capillary blood glucose was markedly improved during the intervention period, with a reduction of mean daily capillary BG from 11.3(8.0, 13.9) mmol/L in the first 24 h to 7.9(6.5,8.9) mmol/L in the last 24 h of the trial (P < 0.001). In addition, the time range below 3.9 mmol/L was decreased from 1.1% to 0.5%.ConclusionsThe clinical decision support system of insulin dosage titration developed using a machine learning algorithm based on the EHRs data was effective and safe in glycemic control in in type 2 diabetic inpatients.Trial registrationsClinicalTrials.gov Identifier: NCT04053959.

A Deep Reinforcement Learning Real-Time Recommendation Model Based on Long and Short-Term Preference

With the development of Internet technology, the problem of information overload has increasingly attracted attention. Nowadays, the recommendation system with excellent performance in information retrieval and filtering would be widely used in the business field. However, most existing recommendation systems are considered a static process, during which recommendations for internet users are often based on pre-trained models. A major disadvantage of these static models is that they are incapable of simulating the interaction process between users and their systems. Moreover, most of these models only consider users’ real-time interests while ignoring their long-term preferences. This paper addresses the abovementioned issues and proposes a new recommendation model, DRR-Max, based on deep reinforcement learning (DRL). In the proposed framework, this paper adopted a state generation module specially designed to obtain users’ long-term and short-term preferences from user profiles and user history score item information. Next, Actor-Critical algorithm is used to simulate the real-time recommendation process.Finally, this paper uses offline and online methods to train the model. In the online mode, the network parameters were dynamically updated to simulate the interaction between the system and users in a real recommendation environment. Experimental results on the two publicly available data sets were used to demonstrate the effectiveness of our proposed model.

Evaluation of an Antifungal Stewardship Initiative Targeting Micafungin at an Academic Medical Center

Delays in the treatment of proven invasive fungal disease have been shown to be harmful. However, empiric treatment for all patients at risk of infection has not demonstrated benefit. This study evaluates the effects of a micafungin stewardship initiative on the duration of therapy and clinical outcomes at the University of Mississippi Medical Center in Jackson, Mississippi. This single-center quasi-experiment evaluated patients who received micafungin. Adult inpatients who received at least one treatment dose of micafungin in the pre-intervention (1 October 2020 to 30 September 2021) or post-intervention (1 October 2021 to 30 April 2022) groups were included. Patients were placed on micafungin for prophylaxis and those who required definitive micafungin therapy were excluded. An algorithm was used to provide real-time recommendations in order to assess change in the treatment days of micafungin therapy. A total of 282 patients were included (141 pre-group versus 141 post-group). Over 80% of the patients included in the study were in an intensive care unit, and other baseline characteristics were similar. The median number of treatment days with micafungin was 4 [IQR 3-6] in the pre-group and 3 [IQR 2-6] in the post-group (p = 0.005). Other endpoints, such as time to discontinuation or de-escalation, hospital mortality, and hospital length of stay, were not significantly different between the groups. An antifungal stewardship initiative can be an effective way to decrease unnecessary empiric antifungal therapy for patients who are at risk of invasive fugal disease.

Modeling the Impact of Route Recommendations in Road Traffic

In this work, effects of real-time route recommendations between two alternative routes are analyzed. The aim is to assess the consequences of route planner applications (apps) on road congestion. For this purpose, the considered model takes the form of a state-dependent switching system that describes a supply-demand mechanism, based on established macroscopic traffic flow models. Through a comprehensive stability analysis of the system, with an emphasis on the dependence on the system parameters (such as road capacities, critical densities, traffic demand), it is shown that real-time recommendations can cause congestion in the network and failure to satisfy user demand.

Comparison of Real-Time and Batch Job Recommendations

Collaborative filtering recommendation systems are traditionally trained in a batch manner and are designed to produce personalized recommendations for a large number of users at the same time. However, in many industrial use cases, it is reasonable to produce recommendations in real time, taking account of very recent user interactions. In this work, we present the implementation of batch and real-time recommendation systems using the example of the RP3Beta model, a simple scalable graph-based model that outperforms multiple more advanced models. Our approach can be utilized by any recommendation system if user-to-item recommendations can be obtained based on item-to-item recommendations. We show that it covers multiple common recommendation models, especially collaborative filtering approaches where user features are not available. We also provide the results of A/B tests comparing these two approaches in a real-world scenario of a job recommendation task, conducted with almost 200,000 OLX users. We report at least 10% more users applying for recommended job ads when using a real-time instead of a batch approach. We believe that our results can help other organizations to take informed decisions about whether to make the effort of moving from a batch to a real-time recommendation setting.

Real-time recommendation method of online education resources based on improved decision tree algorithm

Real-time recommendation method of online education resources based on improved decision tree algorithm

Things of Interest Recommendation with Multidimensional Context Embedding in the Internet of Things

The emerging Internet of Things (IoT) makes users and things closely related together, and the interactions between users and things generate massive context data, where the preference information in time, space, and textual content is embedded. Traditional recommendation methods (e.g., movie, music, and location recommendations) are based on static intrinsic context information, which lacks consideration regarding real-time content and spatiotemporal features, failing to adapt to the personalized recommendation in IoT. Therefore, to meet users’ interests and needs in IoT, a novel effective and efficient recommendation method is urgently needed. The paper focuses on mining users’ things of interest in IoT via leveraging multidimensional context embedding. Specifically, to address the challenge from massive context data embedding different user preference information, the paper employs Convolutional Neural Networks (CNN) to mine the intrinsic content information of things and learn their represent. To solve the real-time recommendation problem, the paper proposes a real-time multimodal model embedded into location, time, and some instant content information to track the features of users and things. Furthermore, the paper proposes a matrix factorization-based framework using the regularization method to fuse real-time context embedding and intrinsic information embedding. The experimental results demonstrate the proposed method tailored to IoT is adaptable and flexible, and able to capture user personalized preference effectively.

A distributed real-time recommender system for big data streams

Recommender Systems (RS) play a crucial role in our lives. As users become continuously connected to the internet, they are less tolerant of obsolete recommendations made by an RS. Online RS has to address three requirements: continuous training and recommendation, handling concept drifts, and the ability to scale. Streaming RS proposed in the literature address the first two requirements only. That is because they run the training process on a single machine.To tackle the third challenge, we propose a Splitting and Replication mechanism for distributed streaming RS. Our mechanism is inspired by the shared-nothing architecture that underpins contemporary big data processing systems. We have applied our mechanism to two well-known approaches for online RS, namely, matrix factorization and item-based collaborative filtering. We conducted experiments comparing the performance with the baseline (single machine). Evaluating different data sets, experiments show online recall improvement by 40% with more than 50% less memory consumption.

Interpretable click-through rate prediction through distillation of the neural additive factorization model

An accurate estimation of the click-through rate (CTR), that is, the probability of clicking on a recommended advertisement item online, is crucial for advertising agencies to make appropriate suggestions to users and maximise their income. However, the existing CTR prediction techniques exhibit two major limitations in estimating accurate CTRs. First, they lack interpretability as they are unable to justify and explain the outcomes of the models. Irrelevant recommendations of products and services may have severe consequences on the personal and financial safety of a user. Second, the high computational complexity of deep-learning models hinders deployment in real-time recommender systems. Thus, this study proposes an interpretable, accurate, and efficient CTR estimator based on the neural additive factorization model (NAFM). The NAFM leverages the model distillation framework, where it effectively learns from a gated ensemble of existing CTR models. The NAFM provides interpretable insights into features and their interactions with a reduced computational cost. The effectiveness and efficiency of the proposed NAFM was verified by applying it on two public click-through datasets, namely, Criteo and Avazu. Unlike post-hoc explainable models, the NAFM avoids an additional phase of training to generate an interpretable model, and its explanations are truthful to the model.

Wingman: A New Movie Recommendation System Algorithm

In this rising new era, the world is growing faster and so are we. People nowadays need effortless services, and recommendation system is one of them. They are present and used everywhere, whether it's advertising, e-commerce or any other domain. Companies such as Netflix, YouTube, Amazon, Spotify rely on recommendation systems to boost their sales. Moreover, it increases the engagement of users with a particular app or website as it tends to show more relevant individualized content. It basically takes the user’s past history as an input and then using filtering techniques it predicts the most suitable item/content for the user as an output. The main objective of this paper is to predict movies most desired by the user with as much accuracy as possible. In this paper, I will provide a brief description of various filtering methods and the particular techniques that I used to achieve the desired result. I have used Tf-Idf to quantify the overview of the movies, then did a pairwise similarity for the Content Based Filtering. This allows the user to watch movies based on their watch history according to the plot of the movies. For Collaborative Filtering, I have used Matrix Factorization to generate real-time recommendations based on user-movie rating matrix. Correlation Coefficient is used to calculate the score. The higher the score, the more likely the movie is liked by the user. I have used the voting system for Hybrid Filtering. Weighted Rating Score is calculated with the help of the votes given to the movies by the users. Then a popularity feature is also used to generate more meaningful movie recommendations. MinMax Scalar is used to normalize the data, and then fifty percent priority is given to both Weighted Rating Score and popularity and a final score is calculated. Using this final score, popular or top movies are generated. The databases used are retrieved from MovieLens and Kaggle. Consequently, my proposed system will generate recommendations based on the plot of the movie using tf-idf and cosine similarity, based on user-movie rating matrix and correlation coefficient, based on voting and popularity basis.

Utilizing data and artificial intelligence to optimize treatment room scheduling and staffing.

436 Background: Tennessee Oncology is a large community oncology practice with over 30 clinics providing 89,000 treatments per year across Tennessee and northern Georgia. Tennessee Oncology’s scheduling application was unable to optimally schedule treatment appointments. This scheduling gap was causing frequent patient delays and employee extended hours. Tennessee Oncology partnered with Smirta, Inc., to develop a data and artificial intelligence (AI) driven scheduling overlay platform that would optimize and simplify cancer treatment scheduling as well as predict scheduling patterns and resource needs. Methods: Named OncoSmart, the scheduling optimization platform ingests historic scheduling data, detailed clinic configuration data including provider and nursing schedules, and available resource data such as treatment room chairs. Utilizing AI, the platform generates optimal scheduling recommendations matching the specific set of services that need to be scheduled. The platform overlays the current scheduling app and provides dynamic, real-time recommendations based on current resource (treatment room, provider, etc.) schedule availabilities and bookings. Tennessee Oncology piloted the scheduling optimization platform at 1 clinic and has currently expanded the pilot to 12 additional clinics. Results: After various ranges of clinic pilot times (6 months to 2 years), Tennessee Oncology treatment volumes have increased by 7%. In parallel to this increase, the optimization platform has helped decrease extended hours by over 32%. The original pilot site has shown major improvement in all 4 primary key performance indicators (KPI): treatment volume +12%; Chair utilization +12%; treatment delay -9%; extended hours -82%. Additionally, using the platform’s predictive analytics capabilities, analyses have been completed to generate optimal treatment scheduling patterns as well as optimal treatment nursing staffing models. Conclusions: Within a short period after deployment, Smirta Inc’s OncoSmart has helped Tennessee Oncology identify better treatment scheduling options for these 13sites. The scheduling optimization platform has proven to be very effective in identifying optimal treatment scheduling strategies and in identifying critical resource bottlenecks. The platform’s clinic management, optimization, nurse assignment, business intelligence, and resource management modules has empowered Tennessee Oncology to better manage critical clinical resources and reduce staff overtime during a period of growth.

A long command subsequence algorithm for manufacturing industry recommendation systems with similarity connection technology

Abstract The manufacturing industry requires a unique recommendation system to suggest products and raw materials, but its performance is often poor in massive data environment. In order to solve the similarity connection problem of large-scale real-time data, the optimised incremental similarity connection method which is used to deal with streaming data can be used to concisely obtain the longest common additive sequence of two given input sequences. This paper, on the basis of the recursion equation, applies a very simple linear space algorithm to solve this problem and adopts new states to carry out similarity connection of incremental data. The experimental results demonstrate that this method can not only ensure the accuracy of real-time recommendation system but also greatly reduce the computed amount.

ПОДХОД К ПОВЫШЕНИЮ ЭФФЕКТИВНОСТИ ПОСТРОЕНИЯ ПЕРСОНАЛИЗИРОВАННЫХ РЕКОМЕНДАЦИЙ В РЕЖИМЕ РЕАЛЬНОГО ВРЕМЕНИ В ОБЛАЧНЫХ СИСТЕМАХ ЭЛЕКТРОННОГО ОБУЧЕНИЯ

The aim of this study is to develop an approach to building recommendations in e-learning systems using graph database algorithms. In the course of the study, the task is to analyze the approach of using classical relational database algorithms in e-learning systems, identify its shortcomings and propose a new approach using graph database algorithms. Analysis, synthesis, comparison and a systematic approach are applied as the research methods. The article proposes a new approach to building real-time recommendations in e-learning systems using graph database algorithms. As the study result, the key advantages of the developed approach are presented. This article shows how easy it is to model a recommendation domain as a graph, using Cypher as the language for getting data from a graph database. Graph databases allow one to instantly respond to user actions and calculate their recommendations in real time, which makes their suggestions more personalized. A scenario from an e-learning system is taken as an application example.

Real-time recommendations for energy-efficient appliance usage in households.

According to several studies, the most influencing factor in a household's energy consumption is user behavior. Changing user behavior to improve energy usage leads to efficient energy consumption, saving money for the consumer and being more friendly for the environment. In this work we propose a framework that aims at assisting households in improving their energy usage by providing real-time recommendations for efficient appliance use. The framework allows for the creation of household-specific and appliance-specific energy consumption profiles by analyzing appliance usage patterns. Based on the household profile and the actual electricity use, real-time recommendations notify users on the appliances that can be switched off in order to reduce consumption. For instance, if a consumer forgets their A/C on at a time that it is usually off (e.g., when there is no one at home), the system will detect this as an outlier and notify the consumer. In the ideal scenario, a household has a smart meter monitoring system installed, that records energy consumption at the appliance level. This is also reflected in the datasets available for evaluating such systems. However, in the general case, the household may only have one main meter reading. In this case, non-intrusive load monitoring (NILM) techniques, which monitor a house's energy consumption using only one meter, and data mining algorithms that disaggregate the consumption into appliance level, can be employed. In this paper, we propose an end-to-end solution to this problem, starting with the energy disaggregation process, and the creation of user profiles that are then fed to the pattern mining and recommendation process, that through an intuitive UI allows users to further refine their energy consumption preferences and set goals. We employ the UK-DALE (UK Domestic Appliance-Level Electricity) dataset for our experimental evaluations and the proof-of-concept implementation. The results show that the proposed framework accurately captures the energy consumption profiles of each household and thus the generated recommendations are matching the actual household energy habits and can help reduce their energy consumption by 2–17%.

Influence of social media in travel decision making

This paper explores the significant impact of social media on travel decision-making processes. With the rapid proliferation of social media platforms, individuals now have access to an extensive array of user-generated content and peer recommendations that greatly influence their travel choices. This study seeks to analyze the ways in which social media platforms shape perceptions, preferences, and decisions related to travel. The research delves into the mechanisms through which social media exerts its influence on travel decisions. It investigates how travelers use platforms like Instagram, Face book, Twitter, and travel-focused review sites to gather information, exchange experiences, and connect with other travelers. Furthermore, the study examines the role of influencers and content creators in shaping travel aspirations and choices. The interactive nature of social media platforms, including the ability to seek real-time advice and engage with travel-related content, adds a layer of immediacy and personalization to the decision-making process. The research on the influence of social media in travel decision making holds significant importance for several key stakeholders, including travelers, the travel industry, and researchers. Understanding the impact of social media on travel decisions empowers travelers to make more informed choices. By recognizing the role of social media, individuals can better navigate the vast array of travel information available and make decisions that align with their preferences and interests. Travelers can use the insights gained from this research to enhance their travel planning processes. They can identify reliable sources of information, seek real-time recommendations, and learn how to effectively engage with content and communities on social media to optimize their travel experiences. Applying the "Weighted Aggregated Sum Product Assessment (WASPAS)" method, those making decisions are able to evaluate possibilities based on a wide range of parameters. It involves providing criteria weights, determining scores for every potential and summing the results to determine the most suitable option. From the result Destination A is in first rank where as Destination E is lower rank.

PupilRec: Leveraging Pupil Morphology for Recommending on Smartphones

As mobile shopping has gradually become the mainstream shopping mode, recommendation systems are gaining an increasingly wide adoption. Existing recommendation systems are mainly based on explicit and implicit user behaviors. However, these user behaviors may not directly indicate users’ inner feelings, causing erroneous user preference estimation and thus leading to inaccurate recommendations. Inspired by our key observation on the correlation between pupil size and users’ inner feelings, we consider using the change of pupil size when browsing to model users’ preferences, so as to achieve targeted recommendations. To this end, we propose PupilRec as a computer-vision-based recommendation framework involving a mobile terminal and a server side. On the mobile terminal, PupilRec collects users’ pupil size change information through the front camera of smartphones; it then preprocesses the raw pupil size data before transmitting them to the server. On the server side, PupilRec utilizes the Tsfresh package and Random Forest algorithm to figure out the key time-series features directly implying user preferences. PupilRec then trains a neural network to fit a user preference model. Using this model, PupilRec predicts user preference to obtain a user–product matrix and further simplifies it by singular value decomposition. Finally, the real-time recommendation is achieved by a collaborative filtering module that retrieves recommended contents to users smartphones. We prototype PupilRec and conduct both experiments and field studies to comprehensively evaluate the effectiveness of PupilRec by recruiting 67 volunteers. The overall results show that PupilRec can accurately estimate users’ preference and can recommend products users interested in.

A Novel Approach for Product Recommendation Using Smartphone Sensor Data

Human Activity-based studies have become an omnipresent research topic in Machine Learning. Considering the countless impacts of human activity on persons' everyday life, we have analyzed the correlation between human activity and their product preferences in our study and proposed that daily human activity could be a metric for product recommendation models. To address this previously unaccounted phenomenon, a new approach is presented in our study that gives real-time recommendations to users by observing their activeness in daily life. However, product recommendation systems mostly believe in ratings, and the purchase behavior of users instead of investigating the precious insights of users' daily activities. But we examined smartphones' GPS sensor data using machine learning algorithms to urge insights from users' daily activeness and proposed a model for predicting the product of interest of the purchasers, based on the activeness of their daily life. Moreover, based on our model, we have introduced a prototype of a real-time recommendation system, especially for the retail shops that rely on users' implicit data from smartphone sensors to form product recommendations. For conducting our study, we developed an android application that—collects embedded smartphone sensor data and can detect objects to provide product recommendations and product details. Experiment shows, that our proffered daily activeness-based recommendation system using smartphone sensor data, performs with a precision of 66%, but it is also a promising performance because it does not use customers' explicit feedback.

Advertising Click-Through Rate Prediction Model Based on an Attention Mechanism and a Neural Network

With the rapid development of the Internet and the rapid expansion of the advertising market, display advertising has become the most popular means of publicity. Accurate advertising recommendation is the guarantee of Internet platform revenue, and accurate advertising click-through rate prediction is the premise of accurate recommendation. According to the requirements of the update rate and real-time of the advertising platform, the coupling relationship of the advertising business module can be divided into two categories: offline business module and online business module. The offline advertising management module is mainly based on the construction of the mathematical model, which is used to mine the complex relationship between users and commodity characteristics; the online advertising management module is mainly based on the real-time feedback of users and changes the recommendation strategy in real time by collecting feedback information. For the offline advertising management module, an advertising click-through rate prediction model based on attention mechanism and neural network is proposed, which is called CAN. This method provides richer feature interaction information for a neural network layer. For the online advertising management module, a real-time recommendation algorithm based on Gaussian process is proposed. This method can solve the universality of specific function assumptions in various environments. Through the coupling of the advertising click-through rate prediction model constructed by the offline advertising management module and the real-time recommendation algorithm of the online advertising management module, the accurate delivery of advertising can be realized.