Quality Of Recommendations Research Articles

FedPOIRec: Privacy-preserving federated poi recommendation with social influence

With the growing number of Location-Based Social Networks, privacy-preserving point-of-interest (POI) recommendation has become a critical challenge when helping users discover potentially interesting new places. Traditional systems take a centralized approach that requires the transmission and collection of private user data. In this work, we present FedPOIRec, a privacy-preserving federated learning approach enhanced with features from user social circles to generate top-N POI recommendations. First, the FedPOIRec framework is built on the principle that local data never leave the owner’s device, while a parameter server blindly aggregates the local updates. Second, the local recommender results are personalized by allowing users to exchange their learned parameters, enabling knowledge transfer among friends. To this end, we propose a privacy-preserving protocol for integrating the preferences of the user’s friends, after the federated computation, by exploiting the properties of the Cheon-Kim-Kim-Song (CKKS) fully homomorphic encryption scheme. To evaluate FedPOIRec, we apply our approach to five real-world datasets using two recommendation models. Extensive experiments demonstrate that FedPOIRec achieves comparable recommendation quality to centralized approaches, while the social integration protocol incurs low computation and communication overhead on the user device.

A comparative study of item space visualizations for recommender systems

Recommender systems aim at supporting users in their search and decision making process by selecting a small number of likely relevant items from a large set of options. Although automatically filtering unmanageably large item sets down to a few recommendations often produces results that match the user’s interests well, it also prevents users from understanding and exploring items in their larger context. This may reduce users’ perception of transparency and controllability of the system. Visualizations have been proposed as a means for overcoming this problem, with some visualizations providing a complete overview of the entire space of available items. However, thus far item space visualizations have rarely been investigated and compared in user studies. To address this, we developed and empirically compared three applications that present the user with personalized music recommendations embedded in a visualization of the entire item space. The three applications display the same item space as a list, as a treemap, and as a map, respectively. We compared these applications in an online user study and found, against our expectations, that they did not differ much in how the recommendations are perceived. Perception of transparency, recommendation quality, and degree of control over the recommendations received relatively high scores over all three applications. However, we did find a difference in hedonic user experience and perceived novelty of the recommendations. Both factors were perceived to be higher in the map condition. Backed up by a mediation analysis, we argue that a halo effect is the reason for the observed perceived novelty: participants transferred the novelty of the application to the novelty of the recommendations.

Application of Improved K-Means Algorithm in Collaborative Recommendation System

With the explosive growth of information resources in the age of big data, mankind has gradually fallen into a serious “information overload” situation. In the face of massive data, collaborative filtering algorithm plays an important role in information filtering and information refinement. However, the recommendation quality and efficiency of collaborative filtering recommendation algorithms are low. The research combines the improved artificial bee colony algorithm with K-means algorithm and applies them to the recommendation system to form a collaborative filtering recommendation algorithm. The experimental results show that the MAE value of the new fitness function is 0.767 on average, which has good separation and compactness in clustering effect. It shows high search accuracy and speed. Compared with the original collaborative filtering algorithm, the average absolute error value of this algorithm is low, and the running time is only 50 s. It improves the recommendation quality and ensures the recommendation efficiency, providing a new research path for the improvement of collaborative filtering recommendation algorithm.

Time-aware Path Reasoning on Knowledge Graph for Recommendation

Reasoning on knowledge graph (KG) has been studied for explainable recommendation due to its ability of providing explicit explanations. However, current KG-based explainable recommendation methods unfortunately ignore the temporal information (such as purchase time, recommend time, etc.), which may result in unsuitable explanations. In this work, we propose a novel Time-aware Path reasoning for Recommendation (TPRec for short) method, which leverages the potential of temporal information to offer better recommendation with plausible explanations. First, we present an efficient time-aware interaction relation extraction component to construct collaborative knowledge graph with time-aware interactions (TCKG for short), and then we introduce a novel time-aware path reasoning method for recommendation. We conduct extensive experiments on three real-world datasets. The results demonstrate that the proposed TPRec could successfully employ TCKG to achieve substantial gains and improve the quality of explainable recommendation.

Analysis of lifestyle modifications and antihypertensive therapy in patients with arterial hypertension and chronic kidney disease (data from the national registry)

The aim is to study the characteristics and quality of drug therapy and recommendations for lifestyle changes in patients with chronic kidney disease and arterial hypertension and, separately, chronic kidney disease and resistant arterial hypertension, observed in primary health care.Materials and methods. The study was carried out on the basis of the AH registry data (N = 43133; 2005-2019 years). Glomerular filtration rate (eGFR) was calculated using the CKD-EPI formula; renal structure and albuminuria were not evaluated. The analysis was performed using the SPSS software (version 22; SPSS Inc).Results. Less than half of all patients reached the target SBP values. ACE inhibitors/ARBs were not prescribed as first-line antihypertensive therapy in 18% of patients with stage 3 CKD. ACE inhibitors were more often prescribed to men than women (70,6% and 66,5%), and ARBs (11,9% and 15,6%). The frequency of prescribing ACE inhibitors, thiazide and thiazide-like diuretics remained practically unchanged at GFR ≥ 60 and 60 ≥ GFR < 30 ml/min/1,73m2, while the frequency of prescribing ARBs, CCBs, loop diuretics increased with a decrease in glomerular filtration rate. Approximately 60% of patients with hypertension were given advice on nutrition and normalization of body weight (among patients with obesity), about 50% — advice on physical activity and ~50% of smokers received advice on smoking cessation. In patients with hypertension and GFR below 60 ml/min/1,73m2, primary care physicians gave advice on lifestyle changes more often than patients with higher GFR. In men with hypertension and 3 stage CKD the incidence of coronary artery disease (2 times), CHF (1,5 times), the incidence of myocardial infarction in history (3,4 times), the incidence of stroke in history (1,9 times) higher than in women. The frequency of the presence of probable resistant hypertension increased up to 23,9% with a decrease of eGFR among patients with uncontrolled hypertension and up to 11% with controlled.Conclusion. For patients with hypertension and CKD, it is necessary to achieve target values of blood pressure, conduct drug therapy aimed at blocking the RAAS (ACE inhibitors /ARBs), select drugs from the CCB group and diuretics as second and third line therapy. In patients with resistant hypertension the addition mineralocorticoid receptor antagonists is necessary.

Knowledge based topic retrieval for recommendations and tourism promotions

The Users regularly share their views on services through internet reviews. Digital tourism allows travelers to plan and manage their trips ahead of time using the internet. The proposed system will enable users to create recommender systems that can help to find new places as per individual interests. The portals such as Tripadvisor and Yelp assist travelers in decision-making and providing feedback. The data available in the form of reviews, blogs, and rating help the researcher build a better recommendation and promotion system. The recommendation system will help tourists discover restaurants often ignored due to a lack of promotion. This research proposes an automated system to perform three subtasks: predicting star ratings from reviews, a feedback model, and a knowledge-based recommender system. The experimental work results show that Random Forest and Decision Tree classifiers give the highest accuracy for predicting star ratings. The research utilizes clustering and topic modeling to identify the topic from the reviews. Moreover, this research uses salience and valence of the topic to generate the feedback model. Furthermore, a recommender system is developed using user preferences, restaurant knowledge, and the valence of each topic of each restaurant. Inverse Document Frequency and K-means clustering have been used to improve the recommendation quality and construct a CSR matrix. This system will link restaurants with similar reviews to ensure that customers find restaurants almost identical to the ones they enjoyed in the past.

Understanding User Preferences in Location-Based Social Networks via a Novel Self-Attention Mechanism

The check-in behaviors of users are ubiquitous in location-based social networks in urban living. Understanding user preferences is critical to improving the recommendation services of social platforms. In addition, great quality of recommendation is also beneficial to sustainable urban living since the user can easily find the point of interest (POI) to visit, which avoids unnecessary consumption, such as a longer time taken for searching or driving. To capture user preferences from their check-in behaviors, advanced methods transform historical records into graph structure data and further leverage graph deep learning-based techniques to learn user preferences. Despite their effectiveness, existing graph deep learning-based methods are limited to the capture of the deep graph’s structural information due to inherent limitations, such as the over-smoothing problem in graph neural networks, further leading to suboptimal performance. To address the above issues, we propose a novel method built on Transformer architecture named spatiotemporal aware transformer (STAT) via a novel graphically aware attention mechanism. In addition, a new temporally aware sampling strategy is developed to reduce the computational cost and enable STAT to deal with large graphs. Extensive experiments on real-world datasets have demonstrated the superiority of the STAT compared to state-of-the-art POI recommendation methods.

Survey on the Objectives of Recommender Systems: Measures, Solutions, Evaluation Methodology, and New Perspectives

Recently, recommender systems have played an increasingly important role in a wide variety of commercial applications to help users find favourite products. Research in the recommender system field has traditionally focused on the accuracy of predictions and the relevance of recommendations. However, other recommendation quality measures may have a significant impact on the overall performance of a recommender system and the satisfaction of users. Hence, researchers’ attention in this field has recently shifted to include other recommender system objectives. This article aims to provide a comprehensive review of recent research efforts on recommender systems based on the objectives achieved: relevance, diversity, novelty, coverage, and serendipity. In addition, the definitions and measures associated with these objectives are reviewed. Furthermore, the article surveys the evaluation methodology used to measure the impact of the main challenges on performance and the new applications of the recommender system. Finally, new perspectives, open issues, and future directions are provided to develop the field.

Recommending Profiles To Social Event Participation In Online Social Network Service Applications Using Deep Learning Techniques

Nowadays, organizing online Social events is growing especially in current pandemic situation due to corona. Thus social network services helps to organize the event oriented online social gathering. The events discussed on social networks can be associated with topics, locations, and time periods. Various literatures have employed clustering mechanism to group the profiles of the user of social networking applications. More research has been carried out in event based online network services as existing models leads to scalability and sparsity problems. Event-based online social networks are used to maintain interest-based groups with high relevancy rate, recommendation quality and predictive accuracy. In order to achieve the above goal, in this paper, we propose a novel framework named as Deep Influence Predict (DIP) which explores the features of Recurrent Neural Network in order identify the target or potential users through different patterns and behaviours of the profile on the social networking service applications. It learns multiple levels of representations and abstractions of the latent data through individual participation record. Further, it extracts the extrinsic and intrinsic properties of the profiles on their social connections and social effects. Specifically, it identifies the distinguishing social groups with different topics and categories as multifaceted interest in iterative process. Finally decision of recommendation for the event is integrated on outcomes of user behaviour model through personnel impact, social relation and equilibrium. Evaluation of the proposed model through various case studies has been implemented using hadoop architecture and validated across various measures such as accuracy on precision, Recall and f measure along scalability and Execution time.

Reinforcement recommendation reasoning through knowledge graphs for explanation path quality

Numerous Knowledge Graphs (KGs) are being created to make Recommender Systems (RSs) not only intelligent but also knowledgeable. Integrating a KG in the recommendation process allows the underlying model to extract reasoning paths between recommended products and already experienced products from the KG. These paths can be leveraged to generate textual explanations to be provided to the user for a given recommendation. However, the existing explainable recommendation approaches based on KG merely optimize the selected reasoning paths for product relevance, without considering any user-level property of the paths for explanation. In this paper, we propose a series of quantitative properties that monitor the quality of the reasoning paths from an explanation perspective, based on recency, popularity, and diversity. We then combine in- and post-processing approaches to optimize for both recommendation quality and reasoning path quality. Experiments on three public data sets show that our approaches significantly increase reasoning path quality according to the proposed properties, while preserving recommendation quality. Source code, data sets, and KGs are available at https://tinyurl.com/bdbfzr4n.

Natural history of ‘silent’ atrial fibrillation from subclinical to asymptomatic: State of the art and need for research

In the last 30 years clinical research in the atrial fibrillation (AF) field has seen a great advance both in terms of productivity and heterogeneity of areas that developed significant new knowledge and evidence. Indeed, it is very easy to verify how in the last 3 decades the number of articles regarding AF registered in MEDLINE has increased exponentially overtime. This large amount of evidence led to significant changes in the guidelines, with an improvement of the overall quality of recommendations and an increase in the high-quality recommendations (despite being still generally low) [ [1] Boriani G. Vitolo M. Lane D.A. Potpara T.S. Lip G.Y. Beyond the 2020 guidelines on atrial fibrillation of the European society of cardiology. Eur J Intern Med. 2021; 86: 1-11https://doi.org/10.1016/j.ejim.2021.01.006 Google Scholar , [2] Barnett A.S. Lewis W.R. Field M.E. Fonarow G.C. Gersh B.J. Page R.L. et al. Quality of evidence underlying the American Heart Association/American College of Cardiology/Heart Rhythm Society Guidelines on the management of atrial fibrillation. JAMA Cardiol. 2017; 2: 319https://doi.org/10.1001/jamacardio.2016.4936 Google Scholar ]. Several areas about AF clinical management are now largely assessed and established, as the need for oral anticoagulant (OAC) prescription, the preference of the non-vitamin K antagonist oral anticoagulants over vitamin k antagonist ones, as well as the assessment of thromboembolic and bleeding risks [ [3] Imberti J.F. Mei D.A. Vitolo M. Bonini N. Proietti M. Potpara T. et al. Comparing atrial fibrillation guidelines: focus on stroke prevention, bleeding risk assessment and oral anticoagulant recommendations. Eur J Intern Med. 2022; 101: 1-7https://doi.org/10.1016/j.ejim.2022.04.023 Google Scholar , [4] Hindricks G. Potpara T. Dagres N. Arbelo E. Bax J.J. Blomström-Lundqvist C. et al. 2020 ESC Guidelines for the diagnosis and management of atrial fibrillation developed in collaboration with the European Association for Cardio-Thoracic Surgery (EACTS). Eur Heart J. 2021; 42: 373-498https://doi.org/10.1093/eurheartj/ehaa612 Google Scholar ]. Also, the need for implementation of integrated care, emerged as the main AF management strategy in the last guidelines, is now largely assessed and evidence based [ 4 Hindricks G. Potpara T. Dagres N. Arbelo E. Bax J.J. Blomström-Lundqvist C. et al. 2020 ESC Guidelines for the diagnosis and management of atrial fibrillation developed in collaboration with the European Association for Cardio-Thoracic Surgery (EACTS). Eur Heart J. 2021; 42: 373-498https://doi.org/10.1093/eurheartj/ehaa612 Google Scholar , 5 Romiti G.F. Pastori D. Rivera-Caravaca J.M. Ding W.Y. Gue Y.X. Menichelli D. et al. Adherence to the “atrial fibrillation better care” pathway in patients with atrial fibrillation: impact on clinical outcomes-a systematic review and meta-analysis of 285,000 patients. Thromb Haemost. 2022; 122: 406-414https://doi.org/10.1055/a-1515-9630 Google Scholar , 6 Romiti G.F. Proietti M. Vitolo M. Bonini N. Fawzy A.M. Ding W.Y. et al. Clinical complexity and impact of the ABC (Atrial fibrillation Better Care) pathway in patients with atrial fibrillation: a report from the ESC-EHRA EURObservational Research Programme in AF General Long-Term Registry. BMC Med. 2022; 20: 326https://doi.org/10.1186/s12916-022-02526-7 Google Scholar ].

Practitioners Versus Users: A Value-Sensitive Evaluation of Current Industrial Recommender System Design

Recommender systems are playing an increasingly important role in alleviating information overload and supporting users' various needs, e.g., consumption, socialization, and entertainment. However, limited research focuses on how values should be extensively considered in industrial deployments of recommender systems, the ignorance of which can be problematic. To fill this gap, in this paper, we adopt Value Sensitive Design to comprehensively explore how practitioners and users recognize different values of current industrial recommender systems. Based on conceptual and empirical investigations, we focus on five values: recommendation quality, privacy, transparency, fairness, and trustworthiness. We further conduct in-depth qualitative interviews with 20 users and 10 practitioners to delve into their opinions about these values. Our results reveal the existence and sources of tensions between practitioners and users in terms of value interpretation, evaluation, and practice, which provide novel implications for designing more human-centric and value-sensitive recommender systems.

Intelligent recommendation system based on decision model of archive translation tasks

How to recruit, test, and train the intelligent recommendation system users, and how to assign the archive translation tasks to all intelligent recommendation system users according to the intelligent matching principles are still a problem that needs to be solved. With the help of proper names and terms in China's Imperial Maritime Customs archives, this manuscript aims to solve the problem. When the corresponding translation, domain or attributes of a proper name or term is known, it will be easier for some archive translation tasks to be completed, and the adaptive archive intelligent recommendation system will also improve the efficiency of intelligent recommendation quality of archive translation tasks. These related domains or attributes are different labels of these archives. To put it simply, multi-label classification means that the same instance can have multiple labels or be labelled into multiple categories, which is called multi-label classification. With the multi-label classification, archives can be classified into different categories, such as the trade archives, preventive archives, personnel archives, etc. The system users are divided into different professional domains by some tests, for instance, system users who are good at economic knowledge and users who have higher language skills. With these labels, the intelligent recommendation system can make the intelligent match between the archives and system users, so as to improve the efficiency and quality of intelligent archive translation tasks. In this manuscript, through multi-label classification, the intelligent recommendation system can realize the intelligent allocation of archive translation tasks to the system users. The intelligent allocation is realized through the construction of intelligent control model, and verifies that the intelligent recommendation system can improve the performance of task allocation over time without the participation of task issuers.

Deep Rating and Review Neural Network for Item Recommendation.

To alleviate the sparsity issue, many recommender systems have been proposed to consider the review text as the auxiliary information to improve the recommendation quality. Despite success, they only use the ratings as the ground truth for error backpropagation. However, the rating information can only indicate the users' overall preference for the items, while the review text contains rich information about the users' preferences and the attributes of the items. In real life, reviews with the same rating may have completely opposite semantic information. If only the ratings are used for error backpropagation, the latent factors of these reviews will tend to be consistent, resulting in the loss of a large amount of review information. In this article, we propose a novel deep model termed deep rating and review neural network (DRRNN) for recommendation. Specifically, compared with the existing models that adopt the review text as the auxiliary information, DRRNN additionally considers both the target rating and target review of the given user-item pair as ground truth for error backpropagation in the training stage. Therefore, we can keep more semantic information of the reviews while making rating predictions. Extensive experiments on four publicly available datasets demonstrate the effectiveness of the proposed DRRNN model in terms of rating prediction.

Med-Tree: A Medical Ontology Tree Combined with the Graph Attention Networks for Medication Recommendation

Medication recommendation based on Electronic Health Records (EHRs) is a significant research direction in the field of intelligent medicine, which aims to recommend personalized medication combinations for patients based on their historical and current physical conditions. However, since the structural and temporal characteristics of medical records are affected by many uncertain factors, there are many limitations in medication recommendation methods based on EHRs. Specifically, most existing works either fail to adequately assess the structural correlation and temporal dependency among various medical entities or ignore existing knowledge of Drug–Drug Interactions (DDI), which could lead to adverse outcomes. These factors contribute to poor recommendation quality. Therefore, we propose a medical ontology tree model combined with the Graph Attention Networks (GAT) for medication recommendations. First, the class hierarchy extracted from the medical ontology and the GAT model is used to learn the ICD-9 codes of diagnoses and procedures, which enriches the semantic representation of medical entities. Secondly, Gate Recurrent Units (GRU) are used to learn the temporal characteristics of medical entities. Finally, memory bank, dynamic memory and DDI graph are used to optimize the hidden layer results, which improve the accuracy of the model. Experimental results show that the proposed model is superior to the previous methods in all evaluation indicators, and the recommended results have a lower DDI rate.

Restorative just culture significantly improves stakeholder inclusion, second victim experiences and quality of recommendations in incident responses

Objective: Matching safety and quality improvements to the complexity of healthcare, Gold Coast Mental Health and Specialist Services implemented a new response to clinical incidents: the Gold Coast Clinical Incident Response Framework (GC-CIRF). It utilises a Restorative Just Culture (RJC) framework and Safety II principles. This paper evaluates its impact.Methods: Staff surveys measured perceptions of just culture and second victim experiences. Quality of recommendations were compared before and after implementation. For the 19 incidents that occurred after the implementation of GC-CIRF, audits of the review processes were undertaken, measuring several components.Results: Results show significant improvement in staff perceptions of just culture and second victim experiences. Review of incident review data showed several shifts in line with Safety II and RJC. The process audit demonstrated inclusion of a broad range of stakeholders, and significant improvements in the quality and strength of recommendations.Conclusions: Embedding RJC and Safety II concepts into the incident review process is associated with improved measures of culture and review outputs. The integration of Safety II concepts and support of cultural shifts will require further work and committed leadership at all levels.

Efficacy and safety of curcuminoids alone in alleviating pain and dysfunction for knee osteoarthritis: a systematic review and meta-analysis of randomized controlled trials

BackgroundCurcuminoids (CURs) are the principal ingredients of Curcuma longa L. [Zingiberaceae] (CL)—an herbal plant used in east Asia to alleviate pain and inflammation. Thus far, the therapeutic effects of CURs for knee osteoarthritis (OA) uncovered by multiple reviews remained uncertain due to broadly involving trials with different agents-combined or CURs-free interventions. Therefore, we formed stringent selection criteria and assessment methods to summarize current evidence on the efficacy and safety of CURs alone in the treatment of knee OA.MethodsA series of databases were searched for randomized controlled trials (RCTs) evaluating the efficacy and safety of CURs for knee OA. Clinical outcomes were evaluated using meta-analysis and the minimum clinically important difference (MCID) for both statistical and clinical significance.ResultsFifteen studies with 1670 patients were included. CURs were significantly more effective than placebo in the improvements of VAS for pain ( WMD: − 1.77, 95% CI: − 2.44 to − 1.09), WOMAC total score ( WMD: − 7.06, 95% CI: − 12.27 to − 1.84), WOMAC pain score ( WMD: − 1.42, 95% CI: − 2.41 to − 0.43), WOMAC function score ( WMD: − 5.04, 95% CI: − 7.65 to − 2.43), and WOMAC stiffness score ( WMD: − 0.54, 95% CI: − 1.03 to − 0.05). Meanwhile, CURs were not inferior to NSAIDs in the improvements of pain- and function-related outcomes. Additionally, CURs did not significantly increase the incidence of adverse events (AEs) compared with placebo ( RR: 1.03, 95% CI: 0.69 to 1.53, P = 0.899, I2 = 23.7%) and NSAIDs (RR: 0.71 0.65, 95% CI: 0.57 0.41 to 0.90 1.03).ConclusionsCURs alone can be expected to achieve considerable analgesic and functional promotion effects for patients with symptomatic knee OA in short term, without inducing an increase of adverse events. However, considering the low quality and substantial heterogeneity of present studies, a cautious and conservative recommendation for broader clinical use of CURs should still be made. Further high-quality studies are necessary to investigate the impact of different dosages, optimization techniques and administration approaches on long-term safety and efficacy of CURs, so as to strengthen clinical decision making for patients with symptomatic knee OA.

Trust-aware location recommendation in location-based social networks: A graph-based approach

With the increase in the use of mobile devices having location-related capabilities, the use of Location-Based Social Networks (LBSN) has also increased, allowing users to share location-embedded information with other users in the social network. By leveraging check-in activities provided by LBSNs, personalized recommendations can be provided. Trust is an important concept in social networks to improve recommendation quality. In this work, we develop a method for predicting the trust scores of LBSN users and propose a trust-aware recommendation technique, TLoRW, to recommend locations to users based on their previous check-ins, the social network, and predicted trust scores of users. In the proposed model, global trust score of user is generated on the basis of check-in history. In addition to trust, spatial context is anther important aspect of TLoRW to generate location recommendations based on the current location of a user. The proposed algorithm runs on a contextual subgraph rather full graph, relaxing the computing resource requirement. We represent a given LBSN with a undirected graph model. Recommendation scores of the locations are generated as the result of the random walk performed on the trust augmented LBSN subgraph. A comprehensive evaluation of TLoRW is conducted by comparing its recommendation performance against baseline techniques, as well as state-of-the-art trust-aware recommendation approaches in the literature, based on benchmark datasets. The experiments reveal that the trust information incorporated into random-walk-based approach improves the accuracy of the recommended locations @5 by minimum 5%.

Dynamic multi-objective sequence-wise recommendation framework via deep reinforcement learning

Sequence-wise recommendation, where recommend exercises to each student step by step, is one of the most exciting tasks in the field of intelligent tutoring systems (ITS). It is important to develop a personalized sequence-wise recommendation framework that immerses students in learning and helps them acquire as much necessary knowledge as possible, rather than merely focusing on providing non-mastered exercises, which is referred to optimize a single objective. However, due to the different knowledge levels of students and the large scale of exercise banks, it is difficult to generate a personalized exercise recommendation for each student. To fully exploit the multifaceted beneficial information collected from e-learning platforms, we design a dynamic multi-objective sequence-wise recommendation framework via deep reinforcement learning, i.e., DMoSwR-DRL, which automatically select the most suitable exercises for each student based on the well-designed domain-objective rewards. Within this framework, the interaction between students and exercises can be explicitly modeled by integrating the actor–critic network and the state representation component, which can greatly help the agent perform effective reinforcement learning. Specifically, we carefully design a state representation module with dynamic recurrent mechanism, which integrates concept information and exercise difficulty level, thus generating a continuous state representation of the student. Subsequently, a flexible reward function is designed to simultaneously optimize the four domain-specific objectives of difficulty, novelty, coverage, and diversity, providing the students with a trade-off sequence-wise recommendation. To set up the online evaluation, we test DMoSwR-DRL on a simulated environment which can model qualitative development of knowledge level and predicts their performance for a given exercise. Comprehensive experiments are conducted on four classical exercise-answer datasets, and the results show the effectiveness and advantages of DMoSwR-DRL in terms of recommendation quality.

Assessment of quality improvement priorities to improve care delivery for patients with ovarian cancer.

307 Background: In the US, more than 12,000 people die from ovarian cancer annually. Although new treatments have recently become available, ovarian cancer care delivery remains suboptimal. The Association of Community Cancer Centers (ACCC) developed a series of initiatives to improve the quality of care for patients with ovarian cancer. In Phase I, ACCC evaluated the needs of cancer programs, conducted 3 quality improvement (QI) workshops, and developed expert-driven Quality of Care recommendations to aid cancer programs in self-assessment and quality improvement. In Phase II, ACCC invited programs to self-assess their level of care using the Quality of Care recommendations. Methods: In 2022, ACCC surveyed 7 ovarian cancer programs across the US. Developed by a multidisciplinary expert steering committee, the 26-question survey assessed the patient populations served, services provided, diagnostic evaluations performed, access to molecular testing, and a self-assessment of quality care delivery and interest in conducting QI projects related to self-identified needs. The survey was administered in Qualtrics and summarized using descriptive statistics. Results: Surveys were completed by 1 community, 2 comprehensive community, 2 hospital-associated, 1 academic comprehensive, and 1 NCI-designated comprehensive cancer programs. The median annual ovarian cases was 40 [range: 2 to 68]. The 7 programs serve diverse racial/ethnic distribution of patients with ovarian cancer ranging from 55% Black to 60% Asian and 30% Pacific Islander. Five programs have a multidisciplinary team for ovarian cancer. All 5 multidisciplinary teams included gynecologic oncology and pathology; 3 teams also included medical oncology and or radiology colleague(s). Sites assessed their level of care for 7 domains of quality care from 1 (lowest) to 5 (highest). The average score was 4.0 for Care Coordination and Patient Education, Treatment Planning, and Disease Surveillance. Diagnosis and Initial Management scored on average 3.8, Equity in Care 3.25, Prevention and Screening 3.25, and Quality of Life 3.2. Sites ranked areas of interest for future QI projects (1 = area of most interest, 7 = area of least interest). The highest ranked domains were Care Coordination and Patient Education as well as Diagnosis and Initial management (average rankings 2.8). These were followed by Quality of Life (average 3.0), Prevention and Screening (Average 3.8), Treatment Planning (Average 4.4), Equity in Care (Average 5.0), and Disease Surveillance (Average 6.2). Conclusions: This assessment will be used to identify 4 additional sites for Phase II QI workshops. Findings from these projects will be combined, summarized, and disseminated to inform future education opportunities and QI initiatives. This initiative aims to help organize and standardize the care of patients with ovarian cancer and improve patient care coordination.