Pengfei Li

Pengfei Li

Auto-generated profile, based on publicly available information

Pengfei Li

Roskilde University, Roskilde, Denmark

Research Focus & Metrics

Academic Career Length

9Years (2018-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
4 July 2018 in IEEE Transactions on Knowledge and Data Engineering
Verified
Journal Article
Closed access
PLPengfei Li
Pengfei Li
HLHua Lu
Hua Lu
NKNattiya Kanhabua
Nattiya Kanhabua
SZSha Zhao
Sha Zhao
GPGang Pan
Gang Pan
As social media users are increasingly going mobile, various location based services (LBS) have been deployed on social media like Twitter. The success of them heavily depends on the availability and accuracy of users' location information. However, only a small fraction of tweets are geo-tagged. Thus, it is necessary to infer locations for tweets in order to attain the purpose of LBS. In this paper, we tackle this problem by scrutinizing Twitter user timelines. First, we split each user's tweet timeline temporally into a number of clusters, each tending to imply a distinct location. Subsequently, we adapt two machine learning models and design classifiers that classify each tweet cluster into one of the pre-defined location classes at the city level. The Bayes based model focuses on the information gain of words with location implications in the user-generated contents. The convolutional LSTM model treats user-generated contents and their associated locations as sequences and employs bidirectional LSTM and convolution operation to infer locations. The two models are evaluated on a large real data set. The results suggest that our models are effective at inferring locations for non-geotagged tweets and the models outperform the state-of-the-art and alternative approaches significantly in terms of inference accuracy.

22 November 2019 in IEEE Transactions on Neural Networks and Learning Systems
Verified
Journal Article
Closed access
WMWenjia Meng
Wenjia Meng
QZQian Zheng
Qian Zheng
LYLong Yang
Long Yang
PLPengfei Li
Pengfei Li
GPGang Pan
Gang Pan
The deep Q-network (DQN) and return-based reinforcement learning are two promising algorithms proposed in recent years. The DQN brings advances to complex sequential decision problems, while return-based algorithms have advantages in making use of sample trajectories. In this brief, we propose a general framework to combine the DQN and most of the return-based reinforcement learning algorithms, named R-DQN. We show that the performance of the traditional DQN can be significantly improved by introducing return-based algorithms. In order to further improve the R-DQN, we design a strategy with two measurements to qualitatively measure the policy discrepancy. We conduct experiments on several representative tasks from the OpenAI Gym and Atari games. The state-of-the-art performance achieved by our method with this proposed strategy validates its effectiveness.

14 August 2019 in IEEE Transactions on Knowledge and Data Engineering
Verified
Journal Article
Closed access
PLPengfei Li
Pengfei Li
HLHua Lu
Hua Lu
QZQian Zheng
Qian Zheng
SLSHIJIAN LI
SHIJIAN LI
GPGang Pan
Gang Pan
Social media users are increasingly going mobile, fostering location based services on social media platforms (e.g., Twitter). Services like friends notification and community detection benefit from co-location judgement that decides if two users are co-located in some point-of-interest (POI). This problem is challenging due to the limited information and the lack of explicit geo-tags in tweets that can be used as labeled data. Our approach is based on a novel concept HisRect features extracted from users' historical visits and recent tweets: The former has impacts on where a user visits in general, whereas the latter gives more hints about where a user is currently. As labeled data is scarce in practice, we design a semi-supervised learning (SSL) framework that leverages unlabeled data to extract HisRect features. Moreover, we employ an embedding neural network layer to process HisRect features of two users, which decides co-location based on the embedding difference between the two features. Our model is evaluated on large sets of real Twitter data from over one million users. The experimental results demonstrate that our HisRect features and SSL framework are highly effective at deciding co-locations. In terms of multiple metrics, our approach clearly outperforms alternative approaches using state-of-the-art techniques.


24 April 2025 in The VLDB Journal
Verified
Journal Article
Open access
PDF file
PLPengfei Li
Pengfei Li
YZYong Zhang
Yong Zhang
WWWenqing Wei
Wenqing Wei
RZRong Zhu
Rong Zhu
BDBolin Ding
Bolin Ding
HLHua Lu
Hua Lu
Approximate query processing (AQP) plays a critical role in modern data analytics. Although machine learning models are used for AQP, existing methods fail to uncover implicit relationships among the underlying data, the aggregate functions in queries, and the query predicates. In this work, we propose a Graph REpresentation Learning-based AQP model (GRELA for short) for answering queries with multiple aggregate functions. GRELA models the aggregate functions and the query predicates as task and clause nodes respectively in a graph and then learns appropriate node representations via its two modules. In particular, the $$\texttt {Encoder}$$ Encoder module coalesces query predicates and underlying data into the representations of clause nodes. The $$\mathbf {\texttt {Graph}}$$ Graph module bridges task nodes and clause nodes such that each task node can aggregate the information from its neighborhood into its representation. Through the inner products of clause and task representations, GRELA is able to make accurate estimates for queries with multiple aggregate functions. Extensive experimental results verify that GRELA outperforms the state-of-the-art AQP methods on different kinds of datasets.