Tensor-based crowdsourced clustering via triangle queries

Vinayak, Ramya Korlakai; Zrnic, Tijana; Hassibi, Babak

doi:10.1109/icassp.2017.7952571

Cited by 3 publications

(9 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Passive Queries Here, we discuss the pros and cons of active querying for crowdsourced clustering when compared to using passive queries. Crowdsourced clustering using passive queries has been previously approached with a two-step process (Gomes et al 2011;Vinayak and Hassibi 2016;Ibrahim and Fu 2021). In the first step, a random or carefully designed pre-determined subset of the n 2 pairs of items, say ⌈r n 2 ⌉ with r ∈ (0, 1] are queried to partially fill a noisy adjacency matrix.…”

Section: Modifications In Queryingmentioning

confidence: 99%

“…We also tried running the active clustering algorithm in (Mazumdar and Saha 2017a), but it failed to run as the initial stage did not yield any clusters even after searching over hyper-parameters. For passive algorithms, we ran the random query algorithm from (Yun and Proutiere 2014), k-means, spectral clustering (McSherry 2001), and convex algorithms (Vinayak and Hassibi 2016). Each algorithm is run 10 times and the results are shown in Table 1, and discussed below:…”

Section: Simulations: Passive Vs Active Queryingmentioning

confidence: 99%

“…A passive strategy for clustering in this scenario is to query all the pairs (i.e., edges), a random subset of pairs (Vinayak and Hassibi 2016), or a specifically constructed subset of pairs (Gomes et al 2011;Ibrahim and Fu 2021) and then perform graph clustering. A major hiccup of such passive strategies is that they can only recover relatively large clusters.…”

Section: Introductionmentioning

confidence: 99%

“…Related Literature: Many prior works that consider the problem of crowdsourced clustering using pairwise similarity queries employ a passive strategy with either a deterministic pattern fixed a priori (Gomes et al 2011;Ibrahim and Fu 2021) or randomly chosen queries (Vinayak, Oymak, and Hassibi 2014;Vinayak and Hassibi 2016). Another related line of work is entity resolution in databases, where the goal is to find data records that represent the same real-world entities.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Crowdsourced Clustering via Active Querying: Practical Algorithm with Theoretical Guarantees

Chen,

Vinayak,

Hassibi

2023

HCOMP

View full text Add to dashboard Cite

We consider the problem of clustering n items into K disjoint clusters using noisy answers from crowdsourced workers to pairwise queries of the type: “Are items i and j from the same cluster?” We propose a novel, practical, simple, and computationally efficient active querying algorithm for crowdsourced clustering. Furthermore, our algorithm does not require knowledge of unknown problem parameters. We show that our algorithm succeeds in recovering the clusters when the crowdworkers provide answers with an error probability less than 1/2 and provide sample complexity bounds on the number of queries made by our algorithm to guarantee successful clustering. While the bounds depend on the error probabilities, the algorithm itself does not require this knowledge. In addition to the theoretical guarantee, we implement and deploy the proposed algorithm on a real crowdsourcing platform to characterize its performance in real-world settings. Based on both the theoretical and the empirical results, we observe that while the total number of queries made by the active clustering algorithm is order-wise better than random querying, the advantage applies most conspicuously when the datasets have small clusters. For datasets with large enough clusters, passive querying can often be more efficient in practice. Our observations and practically implementable active clustering algorithm can inform and aid the design of real-world crowdsourced clustering systems. We make the dataset collected through this work publicly available (and the code to run such experiments).

show abstract

Section: Modifications In Queryingmentioning

confidence: 99%

Section: Simulations: Passive Vs Active Queryingmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Crowdsourced Clustering via Active Querying: Practical Algorithm with Theoretical Guarantees

Chen,

Vinayak,

Hassibi

2023

HCOMP

View full text Add to dashboard Cite

show abstract

“…These groups of overconnected vertices form an essential feature of the structures of most Terrain-Graphs . Their detection is central in a wide variety of fields, such as in biology [ 4 ], in sociology [ 5 ], in linguistics [ 6 ] or in computer sciences [ 7 ], for many tasks as the grouping of most diverse entities [ 8 – 13 ], the pattern detection in data [ 14 ], the prediction of links [ 15 ], the model training [ 16 ], the label assignment [ 17 ], the recommender Algorithms [ 18 ], the data noise removal [ 19 ], or the feature matching [ 20 ].…”

Section: Introductionmentioning

confidence: 99%

Starling: Introducing a mesoscopic scale with Confluence for Graph Clustering

Gaume

2023

PLoS ONE

View full text Add to dashboard Cite

Given a Graph G = (V, E) and two vertices i, j ∈ V, we introduce Confluence(G, i, j), a vertex mesoscopic closeness measure based on short Random walks, which brings together vertices from a same overconnected region of the Graph G, and separates vertices coming from two distinct overconnected regions. Confluence becomes a useful tool for defining a new Clustering quality function QConf(G, Γ) for a given Clustering Γ and for defining a new heuristic Starling to find a partitional Clustering of a Graph G intended to optimize the Clustering quality function QConf. We compare the accuracies of Starling, to the accuracies of three state of the art Graphs Clustering methods: Spectral-Clustering, Louvain, and Infomap. These comparisons are done, on the one hand with artificial Graphs (a) Random Graphs and (b) a classical Graphs Clustering Benchmark, and on the other hand with (c) Terrain-Graphs gathered from real data. We show that with (a), (b) and (c), Starling is always able to obtain equivalent or better accuracies than the three others methods. We show also that with the Benchmark (b), Starling is able to obtain equivalent accuracies and even sometimes better than an Oracle that would only know the expected overconnected regions from the Benchmark, ignoring the concretely constructed edges.

show abstract

Metric Clustering From Triplet Comparisons

Tatli,

Vinayak

2024

2024 60th Annual Allerton Conference on Communication, Control, and Computing

View full text Add to dashboard Cite

Tensor-based crowdsourced clustering via triangle queries

Cited by 3 publications

References 8 publications

Crowdsourced Clustering via Active Querying: Practical Algorithm with Theoretical Guarantees

Crowdsourced Clustering via Active Querying: Practical Algorithm with Theoretical Guarantees

Starling: Introducing a mesoscopic scale with Confluence for Graph Clustering

Metric Clustering From Triplet Comparisons

Contact Info

Product

Resources

About