29 results on '"Lixin Zou"'
Search Results
2. Model-based Unbiased Learning to Rank
- Author
-
Dan Luo, Lixin Zou, Qingyao Ai, Zhiyu Chen, Dawei Yin, and Brian D. Davison
- Subjects
FOS: Computer and information sciences ,Information Retrieval (cs.IR) ,Computer Science - Information Retrieval - Abstract
Unbiased Learning to Rank (ULTR) that learns to rank documents with biased user feedback data is a well-known challenge in information retrieval. Existing methods in unbiased learning to rank typically rely on click modeling or inverse propensity weighting (IPW). Unfortunately, the search engines are faced with severe long-tail query distribution, where neither click modeling nor IPW can handle well. Click modeling suffers from data sparsity problem since the same query-document pair appears limited times on tail queries; IPW suffers from high variance problem since it is highly sensitive to small propensity score values. Therefore, a general debiasing framework that works well under tail queries is in desperate need. To address this problem, we propose a model-based unbiased learning-to-rank framework. Specifically, we develop a general context-aware user simulator to generate pseudo clicks for unobserved ranked lists to train rankers, which addresses the data sparsity problem. In addition, considering the discrepancy between pseudo clicks and actual clicks, we take the observation of a ranked list as the treatment variable and further incorporate inverse propensity weighting with pseudo labels in a doubly robust way. The derived bias and variance indicate that the proposed model-based method is more robust than existing methods. Finally, extensive experiments on benchmark datasets, including simulated datasets and real click logs, demonstrate that the proposed model-based method consistently performs outperforms state-of-the-art methods in various scenarios. The code is available at https://github.com/rowedenny/MULTR., Comment: accepted in WSDM '23; extended version
- Published
- 2023
- Full Text
- View/download PDF
3. User Retention-oriented Recommendation with Decision Transformer
- Author
-
Kesen Zhao, Lixin Zou, Xiangyu Zhao, Maolin Wang, and Dawei Yin
- Subjects
FOS: Computer and information sciences ,Information Retrieval (cs.IR) ,Computer Science - Information Retrieval - Abstract
Improving user retention with reinforcement learning~(RL) has attracted increasing attention due to its significant importance in boosting user engagement. However, training the RL policy from scratch without hurting users' experience is unavoidable due to the requirement of trial-and-error searches. Furthermore, the offline methods, which aim to optimize the policy without online interactions, suffer from the notorious stability problem in value estimation or unbounded variance in counterfactual policy evaluation. To this end, we propose optimizing user retention with Decision Transformer~(DT), which avoids the offline difficulty by translating the RL as an autoregressive problem. However, deploying the DT in recommendation is a non-trivial problem because of the following challenges: (1) deficiency in modeling the numerical reward value; (2) data discrepancy between the policy learning and recommendation generation; (3) unreliable offline performance evaluation. In this work, we, therefore, contribute a series of strategies for tackling the exposed issues. We first articulate an efficient reward prompt by weighted aggregation of meta embeddings for informative reward embedding. Then, we endow a weighted contrastive learning method to solve the discrepancy between training and inference. Furthermore, we design two robust offline metrics to measure user retention. Finally, the significant improvement in the benchmark datasets demonstrates the superiority of the proposed method., Comment: 9 pages, 5 figures
- Published
- 2023
- Full Text
- View/download PDF
4. H-ERNIE
- Author
-
Xiaokai Chu, Jiashu Zhao, Lixin Zou, and Dawei Yin
- Published
- 2022
- Full Text
- View/download PDF
5. Generative Session-based Recommendation
- Author
-
Zhidan Wang, Wenwen Ye, Xu Chen, Wenqiang Zhang, Zhenlei Wang, Lixin Zou, and Weidong Liu
- Published
- 2022
- Full Text
- View/download PDF
6. Fast Semantic Matching via Flexible Contextualized Interaction
- Author
-
Wenwen Ye, Yiding Liu, Lixin Zou, Hengyi Cai, Suqi Cheng, Shuaiqiang Wang, and Dawei Yin
- Published
- 2022
- Full Text
- View/download PDF
7. Subtyping of metastatic breast cancer based on plasma circulating tumor DNA alterations: An observational, multicentre platform study
- Author
-
Zhe-Yu Hu, Yu Tang, Liping Liu, Ning Xie, Can Tian, Binliang Liu, Lixin Zou, Wei Zhou, Yikai Wang, Xuefeng Xia, and Quchang Ouyang
- Subjects
General Medicine - Abstract
After multiple lines of therapies, no guideline or consensus is currently available for the treatment of patients with metastatic breast cancer. This study aims to evaluate the efficacy of a novel re-subtyping and treatment strategy based on ctDNA alterations.This observational, multicentre study recruited 223 patients with metastatic breast cancer intending to receive late-line therapy from Dec 1, 2016, to June 31, 2019. This study took place in Hunan Cancer Hospital, the Forth Hospital of Changsha and Zhuzhou Central Hospital in China. ctDNA alterations were assessed by next-generation sequencing (NGS). Patients with druggable ctDNA alterations were treated with corresponding targeted drugs which are clinically available. Other patients received physician-chosen treatment. This study was registered with ClinicalTrials.gov, NCT05079074.The progression-free survival (hazard ratio: 0.45, 95% Confidence Interval (CI): 0.33-0.62,After multiline traditional pathological HR/HER2 subtype-guided therapies, ctDNA testing could identify druggable ctDNA alterations to guide late-line therapy for patients with metastatic breast cancer.This work was supported by Key Grants of Research and Development in Hunan Province (2018SK2124, 2018SK2120), Natural Science Foundation of Hunan (2019JJ50360), Hunan Provincial Health Commission Project (B2019085, B2019089 and C2019070), and Changsha Science and Technology Project (kq2004125 and kq2004137).
- Published
- 2022
8. Structure prediction of the entire proteome of monkeypox variants
- Author
-
Liangzhen Zheng, Jintao Meng, Mingzhi Lin, Rui Lv, Hongxi Cheng, Lixin Zou, Jinyuan Sun, Linxian Li, Ruobing Ren, and Sheng Wang
- Abstract
Recently, the monkeypox virus has begun to spread in many countries worldwide [1]. The genome sequence of the monkeypox virus variant responsible for the current outbreak has been reported, thus providing an important resource for better understanding the new variant and accelerating vaccine and drug development. Here, we report structure predictions of the whole proteomes of three monkeypox variants, with annotation of potential small-molecule-binding regions of the proteins. Experimentally determined structures with high similarity to monkeypox proteins were vetted through a structure-alignment algorithm. Our work should help accelerate the development of vaccines and drugs.
- Published
- 2022
- Full Text
- View/download PDF
9. Contrastive Disentangled Graph Convolutional Network for Weakly-Supervised Classification
- Author
-
Xiaokai Chu, Jiashu Zhao, Xinxin Fan, Di Yao, Zhihua Zhu, Lixin Zou, Dawei Yin, and Jingping Bi
- Published
- 2022
- Full Text
- View/download PDF
10. Pre-trained Language Model based Ranking in Baidu Search
- Author
-
Zhicong Cheng, Shuaiqiang Wang, Suqi Cheng, Lixin Zou, Dehong Ma, Hengyi Cai, Dawei Yin, Daiting Shi, and Shengqiang Zhang
- Subjects
Search engine ,Information retrieval ,Ranking ,Exploit ,Computer science ,Online search ,Relevance (information retrieval) ,Learning to rank ,Language model ,Latency (engineering) - Abstract
As the heart of a search engine, the ranking system plays a crucial role in satisfying users' information demands. More recently, neural rankers fine-tuned from pre-trained language models (PLMs) establish state-of-the-art ranking effectiveness. However, it is nontrivial to directly apply these PLM-based rankers to the large-scale web search system due to the following challenging issues: (1) the prohibitively expensive computations of massive neural PLMs, especially for long texts in the web document, prohibit their deployments in an online ranking system that demands extremely low latency; (2) the discrepancy between existing ranking-agnostic pre-training objectives and the ad-hoc retrieval scenarios that demand comprehensive relevance modeling is another main barrier for improving the online ranking system; (3) a real-world search engine typically involves a committee of ranking components, and thus the compatibility of the individually fine-tuned ranking model is critical for a cooperative ranking system. In this work, we contribute a series of successfully applied techniques in tackling these exposed issues when deploying the state-of-the-art Chinese pre-trained language model, i.e., ERNIE, in the online search engine system. We first articulate a novel practice to cost-efficiently summarize the web document and contextualize the resultant summary content with the query using a cheap yet powerful Pyramid-ERNIE architecture. Then we endow an innovative paradigm to finely exploit the large-scale noisy and biased post-click behavioral data for relevance-oriented pre-training. We also propose a human-anchored fine-tuning strategy tailored for the online ranking system, aiming to stabilize the ranking signals across various online components. Extensive offline and online experimental results show that the proposed techniques significantly boost the search engine's performance.
- Published
- 2021
- Full Text
- View/download PDF
11. Data-Efficient Reinforcement Learning for Malaria Control
- Author
-
Lixin Zou
- Subjects
Computer science ,business.industry ,Reinforcement learning ,Artificial intelligence ,Machine learning ,computer.software_genre ,Malaria control ,business ,computer - Abstract
Sequential decision-making under cost-sensitive tasks is prohibitively daunting, especially for the problem that has a significant impact on people's daily lives, such as malaria control, treatment recommendation. The main challenge faced by policymakers is to learn a policy from scratch by interacting with a complex environment in a few trials. This work introduces a practical, data-efficient policy learning method, named Variance-Bonus Monte Carlo Tree Search~(VB-MCTS), which can copy with very little data and facilitate learning from scratch in only a few trials. Specifically, the solution is a model-based reinforcement learning method. To avoid model bias, we apply Gaussian Process~(GP) regression to estimate the transitions explicitly. With the GP world model, we propose a variance-bonus reward to measure the uncertainty about the world. Adding the reward to the planning with MCTS can result in more efficient and effective exploration. Furthermore, the derived polynomial sample complexity indicates that VB-MCTS is sample efficient. Finally, outstanding performance on a competitive world-level RL competition and extensive experimental results verify its advantage over the state-of-the-art on the challenging malaria control task.
- Published
- 2021
- Full Text
- View/download PDF
12. Enhanced Doubly Robust Learning for Debiasing Post-Click Conversion Rate Estimation
- Author
-
Hechang Chen, Dawei Yin, Lixin Zou, Wenwen Ye, Shuaiqiang Wang, Yi Chang, Suqi Cheng, Siyuan Guo, and Yiding Liu
- Subjects
FOS: Computer and information sciences ,Selection bias ,Computer Science - Machine Learning ,Computer science ,media_common.quotation_subject ,Estimator ,Variance (accounting) ,Recommender system ,Debiasing ,Machine Learning (cs.LG) ,Computer Science - Information Retrieval ,Robustness (computer science) ,Code (cryptography) ,Imputation (statistics) ,Algorithm ,Information Retrieval (cs.IR) ,media_common - Abstract
Post-click conversion, as a strong signal indicating the user preference, is salutary for building recommender systems. However, accurately estimating the post-click conversion rate (CVR) is challenging due to the selection bias, i.e., the observed clicked events usually happen on users' preferred items. Currently, most existing methods utilize counterfactual learning to debias recommender systems. Among them, the doubly robust (DR) estimator has achieved competitive performance by combining the error imputation based (EIB) estimator and the inverse propensity score (IPS) estimator in a doubly robust way. However, inaccurate error imputation may result in its higher variance than the IPS estimator. Worse still, existing methods typically use simple model-agnostic methods to estimate the imputation error, which are not sufficient to approximate the dynamically changing model-correlated target (i.e., the gradient direction of the prediction model). To solve these problems, we first derive the bias and variance of the DR estimator. Based on it, a more robust doubly robust (MRDR) estimator has been proposed to further reduce its variance while retaining its double robustness. Moreover, we propose a novel double learning approach for the MRDR estimator, which can convert the error imputation into the general CVR estimation. Besides, we empirically verify that the proposed learning scheme can further eliminate the high variance problem of the imputation learning. To evaluate its effectiveness, extensive experiments are conducted on a semi-synthetic dataset and two real-world datasets. The results demonstrate the superiority of the proposed approach over the state-of-the-art methods. The code is available at https://github.com/guosyjlu/MRDR-DL., 10 pages, 3 figures, accepted by SIGIR 2021
- Published
- 2021
- Full Text
- View/download PDF
13. UserSim: User Simulation via Supervised GenerativeAdversarial Network
- Author
-
Xiangyu Zhao, Lixin Zou, Dawei Yin, Jiliang Tang, Long Xia, and Hui Liu
- Subjects
Discriminator ,Computer science ,business.industry ,02 engineering and technology ,Recommender system ,Machine learning ,computer.software_genre ,020204 information systems ,0202 electrical engineering, electronic engineering, information engineering ,Benchmark (computing) ,Reinforcement learning ,020201 artificial intelligence & image processing ,Artificial intelligence ,business ,Generative adversarial network ,computer ,Generator (mathematics) - Abstract
With the recent advances in Reinforcement Learning (RL), there have been tremendous interests in employing RL for recommender systems. However, directly training and evaluating a new RL-based recommendation algorithm needs to collect users’ real-time feedback in the real system, which is time/effort consuming and could negatively impact users’ experiences. Thus, it calls for a user simulator that can mimic real users’ behaviors to pre-train and evaluate new recommendation algorithms. Simulating users’ behaviors in a dynamic system faces immense challenges – (i) the underlying item distribution is complex, and (ii) historical logs for each user are limited. In this paper, we develop a user simulator based on a Generative Adversarial Network (GAN). To be specific, the generator captures the underlying distribution of users’ historical logs and generates realistic logs that can be considered as augmentations of real logs; while the discriminator not only distinguishes real and fake logs but also predicts users’ behaviors. The experimental results based on benchmark datasets demonstrate the effectiveness of the proposed simulator.
- Published
- 2021
- Full Text
- View/download PDF
14. Deep Multifaceted Transformers for Multi-objective Ranking in Large-Scale E-commerce Recommender Systems
- Author
-
Shuaiqiang Wang, Zhuoye Ding, Dawei Yin, Yulong Gu, Lixin Zou, and Yiding Liu
- Subjects
Selection bias ,Training set ,Exploit ,business.industry ,Computer science ,media_common.quotation_subject ,Multi-task learning ,02 engineering and technology ,E-commerce ,Recommender system ,Machine learning ,computer.software_genre ,Single task ,020204 information systems ,0202 electrical engineering, electronic engineering, information engineering ,020201 artificial intelligence & image processing ,Artificial intelligence ,business ,computer ,media_common - Abstract
Recommender Systems have been playing essential roles in e-commerce portals. Existing recommendation algorithms usually learn the ranking scores of items by optimizing a single task (e.g. Click-through rate prediction) based on users' historical click sequences, but they generally pay few attention to simultaneously modeling users' multiple types of behaviors or jointly optimize multiple objectives (e.g. both Click-through rate and Conversion rate), which are both vital for e-commerce sites. In this paper, we argue that it is crucial to formulate users' different interests based on multiple types of behaviors and perform multi-task learning for significant improvement in multiple objectives simultaneously. We propose Deep Multifaceted Transformers (DMT), a novel framework that can model users' multiple types of behavior sequences simultaneously with multiple Transformers. It utilizes Multi-gate Mixture-of-Experts to optimize multiple objectives. Besides, it exploits unbiased learning to reduce the selection bias in the training data. Experiments on JD real production dataset demonstrate the effectiveness of DMT, which significantly outperforms state-of-art methods. DMT has been successfully deployed to serve the main traffic in the commercial Recommender System in JD.com. To facilitate future research, we release the codes and datasets at https://github.com/guyulongcs/CIKM2020_DMT.
- Published
- 2020
- Full Text
- View/download PDF
15. Pseudo Dyna-Q
- Author
-
Weidong Liu, Ting Bai, Long Xia, Pan Du, Dawei Yin, Jian-Yun Nie, Lixin Zou, and Zhuo Zhang
- Subjects
Selection bias ,Computer science ,business.industry ,media_common.quotation_subject ,02 engineering and technology ,Variance (accounting) ,Recommender system ,Machine learning ,computer.software_genre ,020204 information systems ,Convergence (routing) ,Offline learning ,0202 electrical engineering, electronic engineering, information engineering ,Reinforcement learning ,020201 artificial intelligence & image processing ,Artificial intelligence ,business ,Temporal difference learning ,Function (engineering) ,computer ,media_common - Abstract
Applying reinforcement learning (RL) in recommender systems is attractive but costly due to the constraint of the interaction with real customers, where performing online policy learning through interacting with real customers usually harms customer experiences. A practical alternative is to build a recommender agent offline from logged data, whereas directly using logged data offline leads to the problem of selection bias between logging policy and the recommendation policy. The existing direct offline learning algorithms, such as Monte Carlo methods and temporal difference methods are either computationally expensive or unstable on convergence. To address these issues, we propose Pseudo Dyna-Q (PDQ). In PDQ, instead of interacting with real customers, we resort to a customer simulator, referred to as the World Model, which is designed to simulate the environment and handle the selection bias of logged data. During policy improvement, the World Model is constantly updated and optimized adaptively, according to the current recommendation policy. This way, the proposed PDQ not only avoids the instability of convergence and high computation cost of existing approaches but also provides unlimited interactions without involving real customers. Moreover, a proved upper bound of empirical error of reward function guarantees that the learned offline policy has lower bias and variance. Extensive experiments demonstrated the advantages of PDQ on two real-world datasets against state-of-the-arts methods.
- Published
- 2020
- Full Text
- View/download PDF
16. CAMF: Context Aware Matrix Factorization for Social Recommendation
- Author
-
Lixin Zou, Yuan Yao, Yulong Gu, Weidong Liu, and Jiaxing Song
- Subjects
Theoretical computer science ,Artificial Intelligence ,Computer Networks and Communications ,Computer science ,020204 information systems ,0202 electrical engineering, electronic engineering, information engineering ,020201 artificial intelligence & image processing ,Context (language use) ,02 engineering and technology ,Software ,Matrix decomposition - Published
- 2018
- Full Text
- View/download PDF
17. Reinforcement Learning to Optimize Long-term User Engagement in Recommender Systems
- Author
-
Weidong Liu, Jiaxing Song, Dawei Yin, Zhuoye Ding, Long Xia, and Lixin Zou
- Subjects
FOS: Computer and information sciences ,Computer science ,Supervised learning ,02 engineering and technology ,Recommender system ,Computer Science - Information Retrieval ,Term (time) ,User engagement ,Human–computer interaction ,020204 information systems ,0202 electrical engineering, electronic engineering, information engineering ,Reinforcement learning ,020201 artificial intelligence & image processing ,Information Retrieval (cs.IR) - Abstract
Recommender systems play a crucial role in our daily lives. Feed streaming mechanism has been widely used in the recommender system, especially on the mobile Apps. The feed streaming setting provides users the interactive manner of recommendation in never-ending feeds. In such an interactive manner, a good recommender system should pay more attention to user stickiness, which is far beyond classical instant metrics, and typically measured by {\bf long-term user engagement}. Directly optimizing the long-term user engagement is a non-trivial problem, as the learning target is usually not available for conventional supervised learning methods. Though reinforcement learning~(RL) naturally fits the problem of maximizing the long term rewards, applying RL to optimize long-term user engagement is still facing challenges: user behaviors are versatile and difficult to model, which typically consists of both instant feedback~(e.g. clicks, ordering) and delayed feedback~(e.g. dwell time, revisit); in addition, performing effective off-policy learning is still immature, especially when combining bootstrapping and function approximation. To address these issues, in this work, we introduce a reinforcement learning framework --- FeedRec to optimize the long-term user engagement. FeedRec includes two components: 1)~a Q-Network which designed in hierarchical LSTM takes charge of modeling complex user behaviors, and 2)~an S-Network, which simulates the environment, assists the Q-Network and voids the instability of convergence in policy learning. Extensive experiments on synthetic data and a real-world large scale data show that FeedRec effectively optimizes the long-term user engagement and outperforms state-of-the-arts.
- Published
- 2019
- Full Text
- View/download PDF
18. CTRec
- Author
-
Ji-Rong Wen, Lixin Zou, Ting Bai, Weidong Liu, Wayne Xin Zhao, Pan Du, and Jian-Yun Nie
- Subjects
Process (engineering) ,Computer science ,Mechanism (biology) ,Product (category theory) ,Industrial engineering - Abstract
In e-commerce, users' demands are not only conditioned by their profile and preferences, but also by their recent purchases that may generate new demands, as well as periodical demands that depend on purchases made some time ago. We call them respectively short-term demands and long-term demands. In this paper, we propose a novel self-attentive Continuous-Time Recommendation model (CTRec) for capturing the evolving demands of users over time. For modeling such time-sensitive demands, a Demand-aware Hawkes Process (DHP) framework is designed in CTRec to learn from the discrete purchase records of users. More specifically, a convolutional neural network is utilized to capture the short-term demands; and a self-attention mechanism is employed to capture the periodical purchase cycles of long-term demands. All types of demands are fused in DHP to make final continuous-time recommendations. We conduct extensive experiments on four real-world commercial datasets to demonstrate that CTRec is effective for general sequential recommendation problems, including next-item and next-session/basket recommendations. We observe in particular that CTRec is capable of learning the purchase cycles of products and estimating the purchase time of a product given a user.
- Published
- 2019
- Full Text
- View/download PDF
19. Reinforcement Learning to Diversify Top-N Recommendation
- Author
-
Weidong Liu, Dawei Yin, Zhuoye Ding, Lixin Zou, Long Xia, and Jiaxing Song
- Subjects
Artificial neural network ,Computer science ,business.industry ,Maximum coverage problem ,Monte Carlo tree search ,Recommender system ,Machine learning ,computer.software_genre ,Search tree ,Benchmark (computing) ,Reinforcement learning ,Artificial intelligence ,Pruning (decision trees) ,business ,computer - Abstract
In this paper, we study how to recommend both accurate and diverse top-N recommendation, which is a typical instance of the maximum coverage problem. Traditional approaches are to treat the process of constructing the recommendation list as a problem of greedy sequential items selection, which are inevitably sub-optimal. In this paper, we propose a reinforcement learning and neural networks based framework – Diversify top-N Recommendation with Fast Monte Carlo Tree Search (Div-FMCTS) – to optimize the diverse top-N recommendations in a global view. The learning of Div-FMCTS consists of two stages: (1) searching for better recommendation with MCTS; (2) generalizing those plans with the policy and value neural networks. Due to the difficulty of searching over extremely large item permutations, we propose two approaches to speeding up the training process. The first approach is pruning the branches of the search tree by the structure information of the optimal recommendations. The second approach is searching over a randomly chosen small subset of items to quickly harvest the fruits of searching in the generalization with neural networks. Its effectiveness has been proved both empirically and theoretically. Extensive experiments on four benchmark datasets have demonstrated the superiority of Div-FMCTS over state-of-the-art methods.
- Published
- 2019
- Full Text
- View/download PDF
20. HDNN-CF: A hybrid deep neural networks collaborative filtering architecture for event recommendation
- Author
-
Jiaxing Song, Lixin Zou, Weidong Liu, Yuan Yao, and Yulong Gu
- Subjects
Recall ,Artificial neural network ,Event (computing) ,business.industry ,Computer science ,Probabilistic logic ,02 engineering and technology ,010501 environmental sciences ,Semantics ,Machine learning ,computer.software_genre ,01 natural sciences ,Autoencoder ,020204 information systems ,0202 electrical engineering, electronic engineering, information engineering ,Collaborative filtering ,Artificial intelligence ,Representation (mathematics) ,business ,computer ,0105 earth and related environmental sciences - Abstract
Along with the rise of Event-Based Social Networks (EBSNs), event recommendation has become an increasing important problem. However, unlike recommending usual items, such as movies or music, event recommendation suffers from severe cold-start problem, because most events in EBSNs are typically short-lived and registered by only a few users. Additionally, the available feedbacks for events are implicit feedbacks. In this work, we propose a Hybrid Deep Neural Networks Collaborative Filtering Architecture (HDNN-CF) that collaboratively makes use of the events' semantic information and users' implicit feedbacks for event recommendation. Specifically, we extend state-of-the-art method AutoRec to model implicit feedbacks by proposing Probabilistic AutoRec (PAutoRec). We collaboratively train a Stacked Denoise AutoEncoder (SDAE) to learn the deep representation of the semantic information and a PAutoRec to collaborative filter based on implicit feedbacks. Extensive experiments on a real large scale dataset Meetup show that HDNN-CF significantly outperforms state-of-the-art methods by 10% on recall of top 30 recommendations.
- Published
- 2017
- Full Text
- View/download PDF
21. Long short-term memory based recurrent neural networks for collaborative filtering
- Author
-
Yulong Gu, Weidong Liu, Yuan Yao, Lixin Zou, and Jiaxing Song
- Subjects
Ideal (set theory) ,business.industry ,Computer science ,05 social sciences ,Novelty ,020207 software engineering ,02 engineering and technology ,Time step ,Machine learning ,computer.software_genre ,Long short term memory ,Recurrent neural network ,0202 electrical engineering, electronic engineering, information engineering ,Collaborative filtering ,0501 psychology and cognitive sciences ,Artificial intelligence ,business ,computer ,050107 human factors - Abstract
Consuming behaviors of users form sequences ordered by time intuitively. Long Short-Term Memory Based Recurrent Neural Networks(LSTM), which are special kind of Recurrent Neural Networks, are ideal for modeling sequences. In this work, we propose a LSTM based model called CF-LSTM which can model the consuming sequences of users for Collaborative Filtering(CF). To effectively train the CF-LSTM model, we propose the step-combine technique, which processes k ratings at a time step and solves the long sequences problem of ratings. To improve the performance of CF-LSTM, we extend our model with ordinal cost by considering the ordinary nature of users' ratings. Finally, we compare our model with state-of-the-art methods in the metrics of accuracy, novelty and diversity. Extensive experiment results show that CF-LSTM provides highly accurate, novel and diverse recommendations, which outperforms state-of-the-art methods.
- Published
- 2017
- Full Text
- View/download PDF
22. HLGPS: A Home Location Global Positioning System in Location-Based Social Networks
- Author
-
Weidong Liu, Jiaxing Song, Yulong Gu, and Lixin Zou
- Subjects
Personalized search ,Computer science ,business.industry ,020204 information systems ,0202 electrical engineering, electronic engineering, information engineering ,Global Positioning System ,020201 artificial intelligence & image processing ,02 engineering and technology ,Data mining ,computer.software_genre ,business ,computer ,Data modeling - Abstract
The rapid spread of mobile internet and location-acquisition technologies have led to the increasing popularity of Location-Based Social Networks(LBSNs). Users in LBSNs can share their life by checking in at various venues at any time. In LBSNs, identifying home locations of users is significant for effective location-based services like personalized search, targeted advertisement, local recommendation and so on. In this paper, we propose a Home Location Global Positioning System called HLGPS to tackle with the home location identification problem in LBSNs. Firstly, HLGPS uses an influence model named as IME to model edges in LBSNs. Then HLGPS uses a global iteration algorithm based on IME model to position home location of users so that the joint probability of generating all the edges in LBSNs is maximum. Extensive experiments on a large real-world LBSN dataset demonstrate that HLGPS significantly outperforms state-of-the-art methods by 14.7%.
- Published
- 2016
- Full Text
- View/download PDF
23. Context Aware Matrix Factorization for Event Recommendation in Event-Based Social Networks
- Author
-
Jiaxing Song, Yuan Yao, Yulong Gu, Weidong Liu, and Lixin Zou
- Subjects
Context model ,Event (computing) ,Computer science ,business.industry ,Context (language use) ,02 engineering and technology ,Recommender system ,Machine learning ,computer.software_genre ,Popularity ,Matrix decomposition ,020204 information systems ,0202 electrical engineering, electronic engineering, information engineering ,Collaborative filtering ,020201 artificial intelligence & image processing ,Artificial intelligence ,Data mining ,business ,computer - Abstract
Event-based Social Networks(EBSNs) which combine online interactions and offline events among users have experienced increased popularity and rapid growth recently. In EBSNs, event recommendation is significant for users due to the extremely large amount of events. However, the event recommendation problem is rather challenging because it faces a serious cold-start problem: Events have short life time and new events are registered by only a few users. What's more, there are only implicit feedback information. Existing approaches like collaborative filtering methods are not suitable for this scenario. In this paper, we propose a Context Aware Matrix Factorization model called AlphaMF to tackle with the problem. Specifically, AlphaMF is a unified model that combines the Matrix Factorization model which models implicit feedbacks with the Linear contextual features model which models explicit contextual features. Extensive experiments on a large real-world EBSN dataset demonstrate that the AlphaMF model significantly outperforms state-of-the-art methods by 11%.
- Published
- 2016
- Full Text
- View/download PDF
24. Towards Accurate Relation Extraction from Wikipedia
- Author
-
Lixin Zou, Weidong Liu, Yuan Yao, Yulong Gu, and Jiaxing Song
- Subjects
Relation (database) ,business.industry ,Computer science ,Feature extraction ,02 engineering and technology ,computer.software_genre ,Semantics ,Relationship extraction ,Knowledge-based systems ,020204 information systems ,0202 electrical engineering, electronic engineering, information engineering ,Encyclopedia ,020201 artificial intelligence & image processing ,The Internet ,Artificial intelligence ,business ,computer ,Natural language processing - Abstract
Enormous efforts of human volunteers have made Wikipedia become a treasure of textual knowledge. Relation extraction that aims at extracting structured knowledge in the unstructured texts in Wikipedia is an appealing but quite challenging problem because it's hard for machines to understand plain texts. Existing methods are not effective enough because they understand relation types in textual level without exploiting knowledge behind plain texts. In this paper, we propose a novel framework called Athena 2.0 leveraging Semantic Patterns which are patterns that can understand relation types in semantic level to solve this problem. Extensive experiments show that Athena 2.0 significantly outperforms existing methods.
- Published
- 2016
- Full Text
- View/download PDF
25. Phase II study of pazopanib as second-line treatment after sunitinib in patients with metastatic renal cell carcinoma
- Author
-
Qizhan Lin, Mian Xie, Lixin Zou, Chaosheng He, and Jinkun Huang
- Subjects
Oncology ,medicine.medical_specialty ,Second line treatment ,business.industry ,Sunitinib ,Phases of clinical research ,Hematology ,medicine.disease ,Pazopanib ,Renal cell carcinoma ,Internal medicine ,Medicine ,In patient ,business ,medicine.drug - Published
- 2015
- Full Text
- View/download PDF
26. Ant Routing Algorithm for Heavy Load Mobile Ad Hoc Networks
- Author
-
Lixin Zou, Jianli Ding, and Wansheng Tang
- Subjects
Delay-tolerant networking ,Dynamic Source Routing ,Vehicular ad hoc network ,Adaptive quality of service multi-hop routing ,Computer science ,business.industry ,Wireless ad hoc network ,Distributed computing ,ComputerSystemsOrganization_COMPUTER-COMMUNICATIONNETWORKS ,Mobile computing ,Wireless Routing Protocol ,Mobile ad hoc network ,Ad hoc wireless distribution service ,Optimized Link State Routing Protocol ,Link-state routing protocol ,Geocast ,Ad hoc On-Demand Distance Vector Routing ,Mobile wireless sensor network ,Wireless ,Hazy Sighted Link State Routing Protocol ,Destination-Sequenced Distance Vector routing ,business ,ExOR ,Computer network ,Triangular routing - Abstract
A mobile ad hoc network (MANET) is made up of mobile devices that communicate through wireless connections. It has recently attracted much attention in the research and the industry. Routing in mobile ad hoc networks is a challenge problem as a result of highly dynamic topology. This paper describes one Ant Routing method inspired by the behavior of real ants, which can effectively bear network load in Ad-hoc networks. Our simulations in NS-2 show that it performs very well in ad hoc environments, especially in heavy load ad-hoc networks.
- Published
- 2006
- Full Text
- View/download PDF
27. Optic-fiber sensor for steam quality measurement
- Author
-
Pei Liang, Lin Qian, Xinjiang Li, Yang Shen, Lixin Zou, Guirong Zu, and Jianjun Yu
- Subjects
Optical fiber ,Materials science ,business.industry ,Detector ,food and beverages ,complex mixtures ,humanities ,Power (physics) ,law.invention ,Optics ,Surface-area-to-volume ratio ,Fiber optic sensor ,law ,Vapor quality ,business ,Refractive index ,Intensity (heat transfer) - Abstract
The optic-fiber sensor for measuring steam quality has been made out based on the principle of refractivity modulation. It is made up of optic head, conduct optic fiber, light source, detector and electric circuits. Reflectivity on the boundary between optic head and two-phase steam flow can be determined by refractivity of both sides of the boundary. Intensity of reflected light on the boundary is related to the ratio of water and steam. Provided we know the value of temperature or pressure of the two-phase flow, densities of water and vapor can be known and steam quality can be converted from the volume ratio of water and steam. The measuring range of steam quality is 0-100%, and the accuracy is decided by the level of steam quality measured and by resolution power of the sensor. The resolution and sensitivity of the sensor can be tested with two sorts of liquid of known refractive index. Meanwhile, the stability of output of the sensor is also approved. The sensor has been successfully used in measurement of steam quality in a steam-injection oil well with temperature of 270°C and pressure of 8MPa. The measuring result of steam quality tallies with the actual situation.
- Published
- 2005
- Full Text
- View/download PDF
28. Single-particle size measurement by light-scattering method
- Author
-
Lixin Zou, Yongkai Zhao, Longlong Du, Huang Huijie, and Dunwu Lu
- Subjects
Physics ,Optics ,Signal-to-noise ratio ,Particle size measurement ,Optical proximity correction ,business.industry ,Scattering ,Multiangle light scattering ,Optoelectronics ,Curved mirror ,business ,Particle counter ,Light scattering - Abstract
We have designed and constructed a white light optical sensor for optical particle counter (OPC). The optical system of the sensor is a right-angle scattering type optics. It consists of an illumination system and a collection system for scattered light. A large illuminating aperture angle of 24 degrees is obtained by applying an aplanatic singlet in the illumination system, and a large collecting aperture angle of +/- 44 degrees is achieved by a spherical mirror. With the two large apertures, high signal-to-noise ratio and good monotonic light scattering response have been gained.
- Published
- 2000
- Full Text
- View/download PDF
29. High-efficiency optical sensor for optical particle counter
- Author
-
Lixin Zou, Longlong Du, Huang Huijie, and Dunwu Lu
- Subjects
Physics ,Optics ,business.industry ,Optoelectronics ,business ,Particle counter - Published
- 1999
- Full Text
- View/download PDF
Catalog
Discovery Service for Jio Institute Digital Library
For full access to our library's resources, please sign in.