STSiyi Teng

Siyi Teng

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Siyi Teng has a total of 11 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Break Iteration Barrier: Parallelize Priority-Based Graph Processing

    by Siyi Teng, Jeffrey Xu Yu

    Many graph processing systems and graph libraries have been developed to process and analyze graph data efficiently. Among the built-in graph algorithms, priority-based graph algorithms, such as Dijkstra's algorithm and greedy algorithms for combination optimization problems, e.g., influence maximization problem, represent a significant category. They can hardly achieve higher efficiency by parallelism due to their inherent iterative dependencies in the priority queue used. Existing parallelization techniques, e.g., parallel priority queues, struggle to preserve the original processing order of these algorithms. They require larger theoretical time complexities or lose the theoretical guarantees associated with these algorithms. To address these issues, in this paper, we propose PQ + , a priority queue designed to facilitate parallel execution of priority-based graph processing algorithms without altering their inherent process order and preserving the same theoretical time complexity. Extensive experiments on ten real-world datasets across four representative graph processing algorithms validate the effectiveness and efficiency of our proposed approach.

  • Efficient discovery of arbitrary cycles in large-scale networks

    by Siyi Teng, Jeffrey Xu Yu, Jiadong Xie

  • Open Access

    Breaking the Single-Reference-Vector Barrier in Approximate Nearest Neighbor Search

    by Jiadong Xie, Jeffrey Liang, Siyi Teng, Jeffrey Xu Yu, Yingfan Liu

    Approximate nearest neighbor (ANN) searches are commonly employed in various machine learning applications, such as recommendation systems, but traditional ANN searches typically involve only a single reference vector in a query. To broaden the capabilities of ANN search and support multi-reference-vector queries, thereby enabling a wider range of machine learning applications, we introduce all/any-k ANN search. They aim to find vectors that are similar to all or any of the multi-reference vectors in a query, respectively. To effectively and efficiently support all/any-k ANN search, we first propose distance metrics to evaluate the ranking of vectors among those in the dataset for exact all/any-k NN. Building on this, we introduce search algorithms and prove they can search according to the proposed distance metrics on graph indexes designed for traditional ANN. Additionally, we further introduce two-stage search algorithms for all/any-k ANN search to further enhance their search performance. We conduct extensive experiments on real-world datasets to validate the efficiency and effectiveness of our proposed algorithms compared to existing approaches.

  • Open Access

    Beyond Vector Search: Querying With and Without Predicates

    by Jiadong Xie, Jeffrey Xu Yu, Siyi Teng, Yingfan Liu

    k -ANN search has been extensively studied to find k approximate nearest neighbors for a given query vector in a high-dimensional dataset, where a data item is represented as a vector. As there are many new emerging real-world applications that have categorical/numerical attributes associated with vectors, it is highly needed to support k -ANN search with additional predicates on such attributes. In this paper, we study k -ANN queries, q = (v q , c q ), where v q is a query vector and c q is a predicate on categorical/numerical attributes. Note that the conventional k -ANN search is a k -ANN query when c q = ∅. In the literature, some can support the cases when c q = ∅, some can support the cases when c q is on categorical attributes, and some can support the cases when c q is on numerical attributes. But none of them can support all cases efficiently. In this paper, we propose an all-in-one approach. Our approach supports conventional k -ANN search in the same way as the state-of-the-art approaches, and supports the predicates in a similar or even better way compared to the approaches that are tailored to support either categorical attributes or numerical attributes. We conduct extensive performance studies and confirm the accuracy and the efficiency of our approach in comparison with the state-of-the-art approaches.


  • Open Access

    Optimizing Network Resilience via Vertex Anchoring

    by Siyi Teng, Jiadong Xie, Fan Zhang, Can Lu, Juntao Fang, Kai Wang

    Network resilience is a critical ability of a network to maintain its functionality against disturbances. A network is resilient/robust when a large portion of the nodes are to be better engaged in the network, i.e., they are less likely to leave given the changes on the network. Existing studies validate that the engagement of a node can be well captured by its coreness on network topology. Therefore, it is promising to maximize the number of nodes with increasing coreness values. In this paper, we propose and study thefollower maximization problem: maximizing the resilience gain (the number of coreness-increased vertices) via anchoring a set of vertices within a given budget. We prove that the problem is NP-hard and W[2]-hard, and it is NP-hard to approximate within an O(n^1-ε ) factor. We first propose an advanced greedy approach, followed by a time-dependent framework designed to quickly find high-quality results. The framework is initialized by the advanced greedy algorithm and incorporates novel techniques for optimizing the search space. The effectiveness and efficiency of our solution are verified with extensive experiments on 8 real-life datasets. Our source codes are available at https://github.com/Tsyxxxka/Follower-Maximization.


  • Open Access

    IMinimize: A System for Negative Influence Minimization via Vertex Blocking

    by Siyi Teng, Jiadong Xie, Mingkai Zhang, Kai Wang, Fan Zhang

    The rapid rise and prevalence of social platforms have created great demands on effective schemes to limit the influence of negative information, e.g., blocking key vertices for influence minimization. However, there is currently no system providing practical schemes to solve the negative influence minimization problem with a blocking budget effectively and efficiently in the literature. In this demo, we present IMinimize, the first interactive system that provides audiences with vertex-blocking schemes over different budgets and demonstrates via visualization for comparison vividly and directly, aiming to help minimize the negative influence spreading in networks. Our IMinimize system applies an advanced greedy algorithm to select blocked vertices with both high efficiency and effectiveness. Furthermore, we extend IMinimize to the application of epidemic controlling and prevention and show the usability of IMinimize through two case studies of real-life applications.