Matteo Lissandrini, D. Mottin, Themis Palpanas, Yannis Velegrakis
We consider the task of exploratory search through graph queries on knowledge graphs. We propose to assist the user by expanding the query with intuitive suggestions to provide a more informative (full) query that can retrieve more detailed and relevant answers. To achieve this result, we propose a model that can bridge graph search paradigms with well-established techniques for information-retrieval. Our approach does not require any additional knowledge from the user and builds on principled language modelling approaches. We empirically show the effectiveness and efficiency of our approach on a large knowledge graph and how our suggestions are able to help build more complete and informative queries.
{"title":"Graph-Query Suggestions for Knowledge Graph Exploration","authors":"Matteo Lissandrini, D. Mottin, Themis Palpanas, Yannis Velegrakis","doi":"10.1145/3366423.3380005","DOIUrl":"https://doi.org/10.1145/3366423.3380005","url":null,"abstract":"We consider the task of exploratory search through graph queries on knowledge graphs. We propose to assist the user by expanding the query with intuitive suggestions to provide a more informative (full) query that can retrieve more detailed and relevant answers. To achieve this result, we propose a model that can bridge graph search paradigms with well-established techniques for information-retrieval. Our approach does not require any additional knowledge from the user and builds on principled language modelling approaches. We empirically show the effectiveness and efficiency of our approach on a large knowledge graph and how our suggestions are able to help build more complete and informative queries.","PeriodicalId":20754,"journal":{"name":"Proceedings of The Web Conference 2020","volume":"114 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2020-04-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"77693772","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Weichao Wang, Shi Feng, Wei Gao, Daling Wang, Yifei Zhang
In open-domain dialogue systems, dialogue cues such as emotion, persona, and emoji can be incorporated into conversation models for strengthening the semantic relevance of generated responses. Existing neural response generation models either incorporate dialogue cue into decoder’s initial state or embed the cue indiscriminately into the state of every generated word, which may cause the gradients of the embedded cue to vanish or disturb the semantic relevance of generated words during back propagation. In this paper, we propose a Cue Adaptive Decoder (CueAD) that aims to dynamically determine the involvement of a cue at each generation step in the decoding. For this purpose, we extend the Gated Recurrent Unit (GRU) network with an adaptive cue representation for facilitating cue incorporation, in which an adaptive gating unit is utilized to decide when to incorporate cue information so that the cue can provide useful clues for enhancing the semantic relevance of the generated words. Experimental results show that CueAD outperforms state-of-the-art baselines with large margins.
{"title":"A Cue Adaptive Decoder for Controllable Neural Response Generation","authors":"Weichao Wang, Shi Feng, Wei Gao, Daling Wang, Yifei Zhang","doi":"10.1145/3366423.3380008","DOIUrl":"https://doi.org/10.1145/3366423.3380008","url":null,"abstract":"In open-domain dialogue systems, dialogue cues such as emotion, persona, and emoji can be incorporated into conversation models for strengthening the semantic relevance of generated responses. Existing neural response generation models either incorporate dialogue cue into decoder’s initial state or embed the cue indiscriminately into the state of every generated word, which may cause the gradients of the embedded cue to vanish or disturb the semantic relevance of generated words during back propagation. In this paper, we propose a Cue Adaptive Decoder (CueAD) that aims to dynamically determine the involvement of a cue at each generation step in the decoding. For this purpose, we extend the Gated Recurrent Unit (GRU) network with an adaptive cue representation for facilitating cue incorporation, in which an adaptive gating unit is utilized to decide when to incorporate cue information so that the cue can provide useful clues for enhancing the semantic relevance of the generated words. Experimental results show that CueAD outperforms state-of-the-art baselines with large margins.","PeriodicalId":20754,"journal":{"name":"Proceedings of The Web Conference 2020","volume":"73 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2020-04-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84395133","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Clique and near-clique counts are important graph properties with applications in graph generation, graph modeling, graph analytics, community detection among others. They are the archetypal examples of dense subgraphs. While there are several different definitions of near-cliques, most of them share the attribute that they are cliques that are missing a small number of edges. Clique counting is itself considered a challenging problem. Counting near-cliques is significantly harder more so since the search space for near-cliques is orders of magnitude larger than that of cliques. We give a formulation of a near-clique as a clique that is missing a constant number of edges. We exploit the fact that a near-clique contains a smaller clique, and use techniques for clique sampling to count near-cliques. This method allows us to count near-cliques with 1 or 2 missing edges, in graphs with tens of millions of edges. To the best of our knowledge, there was no known efficient method for this problem, and we obtain a 10x − 100x speedup over existing algorithms for counting near-cliques. Our main technique is a space efficient adaptation of the Turán Shadow sampling approach, recently introduced by Jain and Seshadhri (WWW 2017). This approach constructs a large recursion tree (called the Turán Shadow) that represents cliques in a graph. We design a novel algorithm that builds an estimator for near-cliques, using a online, compact construction of the Turán Shadow.
{"title":"Provably and Efficiently Approximating Near-cliques using the Turán Shadow: PEANUTS","authors":"Shweta Jain, C. Seshadhri","doi":"10.1145/3366423.3380264","DOIUrl":"https://doi.org/10.1145/3366423.3380264","url":null,"abstract":"Clique and near-clique counts are important graph properties with applications in graph generation, graph modeling, graph analytics, community detection among others. They are the archetypal examples of dense subgraphs. While there are several different definitions of near-cliques, most of them share the attribute that they are cliques that are missing a small number of edges. Clique counting is itself considered a challenging problem. Counting near-cliques is significantly harder more so since the search space for near-cliques is orders of magnitude larger than that of cliques. We give a formulation of a near-clique as a clique that is missing a constant number of edges. We exploit the fact that a near-clique contains a smaller clique, and use techniques for clique sampling to count near-cliques. This method allows us to count near-cliques with 1 or 2 missing edges, in graphs with tens of millions of edges. To the best of our knowledge, there was no known efficient method for this problem, and we obtain a 10x − 100x speedup over existing algorithms for counting near-cliques. Our main technique is a space efficient adaptation of the Turán Shadow sampling approach, recently introduced by Jain and Seshadhri (WWW 2017). This approach constructs a large recursion tree (called the Turán Shadow) that represents cliques in a graph. We design a novel algorithm that builds an estimator for near-cliques, using a online, compact construction of the Turán Shadow.","PeriodicalId":20754,"journal":{"name":"Proceedings of The Web Conference 2020","volume":"26 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2020-04-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85872854","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Understanding representativeness in cellular web logs at city scale is essential for web applications. Most of the existing work on cellular web analyses or applications is built upon data from a single network in a city, which may not be representative of the overall usage patterns since multiple cellular networks coexist in most cities in the world. In this paper, we conduct the first comprehensive investigation of multiple cellular networks in a city with a 100% user penetration rate. We study web usage pattern (e.g., internet access services) correlation and difference between diverse cellular networks in terms of spatial and temporal dimensions to quantify the representativeness of web usage from a single network in usage patterns of all users in the same city. Moreover, relying on three external datasets, we study the correlation between the representativeness and contextual factors (e.g., Point-of-Interest, population, and mobility) to explain the potential causalities for the representativeness difference. We found that contextual diversity is a key reason for representativeness difference, and representativeness has a significant impact on the performance of real-world applications. Based on the analysis results, we further design a correction model to address the bias of single cellphone networks and improve representativeness by 45.8%.
{"title":"CellRep: Usage Representativeness Modeling and Correction Based on Multiple City-Scale Cellular Networks","authors":"Zhihan Fang, Guang Wang, Shuai Wang, Chaoji Zuo, Fan Zhang, Desheng Zhang","doi":"10.1145/3366423.3380141","DOIUrl":"https://doi.org/10.1145/3366423.3380141","url":null,"abstract":"Understanding representativeness in cellular web logs at city scale is essential for web applications. Most of the existing work on cellular web analyses or applications is built upon data from a single network in a city, which may not be representative of the overall usage patterns since multiple cellular networks coexist in most cities in the world. In this paper, we conduct the first comprehensive investigation of multiple cellular networks in a city with a 100% user penetration rate. We study web usage pattern (e.g., internet access services) correlation and difference between diverse cellular networks in terms of spatial and temporal dimensions to quantify the representativeness of web usage from a single network in usage patterns of all users in the same city. Moreover, relying on three external datasets, we study the correlation between the representativeness and contextual factors (e.g., Point-of-Interest, population, and mobility) to explain the potential causalities for the representativeness difference. We found that contextual diversity is a key reason for representativeness difference, and representativeness has a significant impact on the performance of real-world applications. Based on the analysis results, we further design a correction model to address the bias of single cellphone networks and improve representativeness by 45.8%.","PeriodicalId":20754,"journal":{"name":"Proceedings of The Web Conference 2020","volume":"26 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2020-04-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"88231172","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Spatial event forecasting is challenging and crucial for urban sensing scenarios, which is beneficial for a wide spectrum of spatial-temporal mining applications, ranging from traffic management, public safety, to environment policy making. In spite of significant progress has been made to solve spatial-temporal prediction problem, most existing deep learning based methods based on a coarse-grained spatial setting and the success of such methods largely relies on data sufficiency. In many real-world applications, predicting events with a fine-grained spatial resolution do play a critical role to provide high discernibility of spatial-temporal data distributions. However, in such cases, applying existing methods will result in weak performance since they may not well capture the quality spatial-temporal representations when training triple instances are highly imbalanced across locations and time. To tackle this challenge, we develop a hierarchically structured Spatial-Temporal ransformer network (STtrans) which leverages a main embedding space to capture the inter-dependencies across time and space for alleviating the data imbalance issue. In our STtrans framework, the first-stage transformer module discriminates different types of region and time-wise relations. To make the latent spatial-temporal representations be reflective of the relational structure between categories, we further develop a cross-category fusion transformer network to endow STtrans with the capability to preserve the semantic signals in a fully dynamic manner. Finally, an adversarial training strategy is introduced to yield a robust spatial-temporal learning under data imbalance. Extensive experiments on real-world imbalanced spatial-temporal datasets from NYC and Chicago demonstrate the superiority of our method over various state-of-the-art baselines.
{"title":"Hierarchically Structured Transformer Networks for Fine-Grained Spatial Event Forecasting","authors":"Xian Wu, Chao Huang, Chuxu Zhang, N. Chawla","doi":"10.1145/3366423.3380296","DOIUrl":"https://doi.org/10.1145/3366423.3380296","url":null,"abstract":"Spatial event forecasting is challenging and crucial for urban sensing scenarios, which is beneficial for a wide spectrum of spatial-temporal mining applications, ranging from traffic management, public safety, to environment policy making. In spite of significant progress has been made to solve spatial-temporal prediction problem, most existing deep learning based methods based on a coarse-grained spatial setting and the success of such methods largely relies on data sufficiency. In many real-world applications, predicting events with a fine-grained spatial resolution do play a critical role to provide high discernibility of spatial-temporal data distributions. However, in such cases, applying existing methods will result in weak performance since they may not well capture the quality spatial-temporal representations when training triple instances are highly imbalanced across locations and time. To tackle this challenge, we develop a hierarchically structured Spatial-Temporal ransformer network (STtrans) which leverages a main embedding space to capture the inter-dependencies across time and space for alleviating the data imbalance issue. In our STtrans framework, the first-stage transformer module discriminates different types of region and time-wise relations. To make the latent spatial-temporal representations be reflective of the relational structure between categories, we further develop a cross-category fusion transformer network to endow STtrans with the capability to preserve the semantic signals in a fully dynamic manner. Finally, an adversarial training strategy is introduced to yield a robust spatial-temporal learning under data imbalance. Extensive experiments on real-world imbalanced spatial-temporal datasets from NYC and Chicago demonstrate the superiority of our method over various state-of-the-art baselines.","PeriodicalId":20754,"journal":{"name":"Proceedings of The Web Conference 2020","volume":"18 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2020-04-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79670316","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Ramnath Kumar, S. Yadav, Raminta Daniulaityte, Francois R. Lamy, K. Thirunarayan, Usha Lokala, A. Sheth
Darknet crypto markets are online marketplaces using crypto currencies (e.g., Bitcoin, Monero) and advanced encryption techniques to offer anonymity to vendors and consumers trading for illegal goods or services. The exact volume of substances advertised and sold through these crypto markets is difficult to assess, at least partially, because vendors tend to maintain multiple accounts (or Sybil accounts) within and across different crypto markets. Linking these different accounts will allow us to accurately evaluate the volume of substances advertised across the different crypto markets by each vendor. In this paper, we present a multi-view unsupervised framework (eDarkFind) that helps modeling vendor characteristics and facilitates Sybil account detection. We employ a multi-view learning paradigm to generalize and improve the performance by exploiting the diverse views from multiple rich sources such as BERT, stylometric, and location representation. Our model is further tailored to take advantage of domain-specific knowledge such as the Drug Abuse Ontology to take into consideration the substance information. We performed extensive experiments and demonstrated that the multiple views obtained from diverse sources can be effective in linking Sybil accounts. Our proposed eDarkFind model achieves an accuracy of 98% on three real-world datasets which shows the generality of the approach.
{"title":"eDarkFind: Unsupervised Multi-view Learning for Sybil Account Detection","authors":"Ramnath Kumar, S. Yadav, Raminta Daniulaityte, Francois R. Lamy, K. Thirunarayan, Usha Lokala, A. Sheth","doi":"10.1145/3366423.3380263","DOIUrl":"https://doi.org/10.1145/3366423.3380263","url":null,"abstract":"Darknet crypto markets are online marketplaces using crypto currencies (e.g., Bitcoin, Monero) and advanced encryption techniques to offer anonymity to vendors and consumers trading for illegal goods or services. The exact volume of substances advertised and sold through these crypto markets is difficult to assess, at least partially, because vendors tend to maintain multiple accounts (or Sybil accounts) within and across different crypto markets. Linking these different accounts will allow us to accurately evaluate the volume of substances advertised across the different crypto markets by each vendor. In this paper, we present a multi-view unsupervised framework (eDarkFind) that helps modeling vendor characteristics and facilitates Sybil account detection. We employ a multi-view learning paradigm to generalize and improve the performance by exploiting the diverse views from multiple rich sources such as BERT, stylometric, and location representation. Our model is further tailored to take advantage of domain-specific knowledge such as the Drug Abuse Ontology to take into consideration the substance information. We performed extensive experiments and demonstrated that the multiple views obtained from diverse sources can be effective in linking Sybil accounts. Our proposed eDarkFind model achieves an accuracy of 98% on three real-world datasets which shows the generality of the approach.","PeriodicalId":20754,"journal":{"name":"Proceedings of The Web Conference 2020","volume":"44 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2020-04-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83645316","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Attention check questions have become commonly used in online surveys published on popular crowdsourcing platforms as a key mechanism to filter out inattentive respondents and improve data quality. However, little research considers the vulnerabilities of this important quality control mechanism that can allow attackers including irresponsible and malicious respondents to automatically answer attention check questions for efficiently achieving their goals. In this paper, we perform the first study to investigate such vulnerabilities, and demonstrate that attackers can leverage deep learning techniques to pass attention check questions automatically. We propose AC-EasyPass, an attack framework with a concrete model, that combines convolutional neural network and weighted feature reconstruction to easily pass attention check questions. We construct the first attention check question dataset that consists of both original and augmented questions, and demonstrate the effectiveness of AC-EasyPass. We explore two simple defense methods, adding adversarial sentences and adding typos, for survey designers to mitigate the risks posed by AC-EasyPass; however, these methods are fragile due to their limitations from both technical and usability perspectives, underlining the challenging nature of defense. We hope our work will raise sufficient attention of the research community towards developing more robust attention check mechanisms. More broadly, our work intends to prompt the research community to seriously consider the emerging risks posed by the malicious use of machine learning techniques to the quality, validity, and trustworthiness of crowdsourcing and social computing.
{"title":"Attention Please: Your Attention Check Questions in Survey Studies Can Be Automatically Answered","authors":"Weiping Pei, Arthur Mayer, Kaylynn Tu, Chuan Yue","doi":"10.1145/3366423.3380195","DOIUrl":"https://doi.org/10.1145/3366423.3380195","url":null,"abstract":"Attention check questions have become commonly used in online surveys published on popular crowdsourcing platforms as a key mechanism to filter out inattentive respondents and improve data quality. However, little research considers the vulnerabilities of this important quality control mechanism that can allow attackers including irresponsible and malicious respondents to automatically answer attention check questions for efficiently achieving their goals. In this paper, we perform the first study to investigate such vulnerabilities, and demonstrate that attackers can leverage deep learning techniques to pass attention check questions automatically. We propose AC-EasyPass, an attack framework with a concrete model, that combines convolutional neural network and weighted feature reconstruction to easily pass attention check questions. We construct the first attention check question dataset that consists of both original and augmented questions, and demonstrate the effectiveness of AC-EasyPass. We explore two simple defense methods, adding adversarial sentences and adding typos, for survey designers to mitigate the risks posed by AC-EasyPass; however, these methods are fragile due to their limitations from both technical and usability perspectives, underlining the challenging nature of defense. We hope our work will raise sufficient attention of the research community towards developing more robust attention check mechanisms. More broadly, our work intends to prompt the research community to seriously consider the emerging risks posed by the malicious use of machine learning techniques to the quality, validity, and trustworthiness of crowdsourcing and social computing.","PeriodicalId":20754,"journal":{"name":"Proceedings of The Web Conference 2020","volume":"40 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2020-04-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89664787","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
R. Guerraoui, Anne-Marie Kermarrec, Olivier Ruas, François Taïani
We propose GoldFinger, a new compact and fast-to-compute binary representation of datasets to approximate Jaccard’s index. We illustrate the effectiveness of GoldFinger on the emblematic big data problem of K-Nearest-Neighbor (KNN) graph construction and show that GoldFinger can drastically accelerate a large range of existing KNN algorithms with little to no overhead. As a side effect, we also show that the compact representation of the data protects users’ privacy for free by providing k-anonymity and l-diversity. Our extensive evaluation of the resulting approach on several realistic datasets shows that our approach delivers speedups of up to 78.9% compared to the use of raw data while only incurring a negligible to moderate loss in terms of KNN quality. To convey the practical value of such a scheme, we apply it to item recommendation and show that the loss in recommendation quality is negligible.
{"title":"Smaller, Faster & Lighter KNN Graph Constructions","authors":"R. Guerraoui, Anne-Marie Kermarrec, Olivier Ruas, François Taïani","doi":"10.1145/3366423.3380184","DOIUrl":"https://doi.org/10.1145/3366423.3380184","url":null,"abstract":"We propose GoldFinger, a new compact and fast-to-compute binary representation of datasets to approximate Jaccard’s index. We illustrate the effectiveness of GoldFinger on the emblematic big data problem of K-Nearest-Neighbor (KNN) graph construction and show that GoldFinger can drastically accelerate a large range of existing KNN algorithms with little to no overhead. As a side effect, we also show that the compact representation of the data protects users’ privacy for free by providing k-anonymity and l-diversity. Our extensive evaluation of the resulting approach on several realistic datasets shows that our approach delivers speedups of up to 78.9% compared to the use of raw data while only incurring a negligible to moderate loss in terms of KNN quality. To convey the practical value of such a scheme, we apply it to item recommendation and show that the loss in recommendation quality is negligible.","PeriodicalId":20754,"journal":{"name":"Proceedings of The Web Conference 2020","volume":"75 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2020-04-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"77915239","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Kasper Green Larsen, M. Mitzenmacher, Charalampos E. Tsourakakis
Clustering, i.e., finding groups in the data, is a problem that permeates multiple fields of science and engineering. Recently, the problem of clustering with a noisy oracle has drawn attention due to various applications including crowdsourced entity resolution [33], and predicting signs of interactions in large-scale online social networks [20, 21]. Here, we consider the following fundamental model for two clusters as proposed by Mitzenmacher and Tsourakakis [28], and Mazumdar and Saha [25]; there exist n items, belonging to two unknown groups. We are allowed to query any pair of nodes whether they belong to the same cluster or not, but the answer to the query is corrupted with some probability . Let 1 > δ = 1 − 2q > 0 be the bias. In this work, we provide a polynomial time algorithm that recovers all signs correctly with high probability in the presence of noise with queries. This is the best known result for this problem for all but tiny δ, improving on the current state-of-the-art due to Mazumdar and Saha [25].
{"title":"Clustering with a faulty oracle","authors":"Kasper Green Larsen, M. Mitzenmacher, Charalampos E. Tsourakakis","doi":"10.1145/3366423.3380045","DOIUrl":"https://doi.org/10.1145/3366423.3380045","url":null,"abstract":"Clustering, i.e., finding groups in the data, is a problem that permeates multiple fields of science and engineering. Recently, the problem of clustering with a noisy oracle has drawn attention due to various applications including crowdsourced entity resolution [33], and predicting signs of interactions in large-scale online social networks [20, 21]. Here, we consider the following fundamental model for two clusters as proposed by Mitzenmacher and Tsourakakis [28], and Mazumdar and Saha [25]; there exist n items, belonging to two unknown groups. We are allowed to query any pair of nodes whether they belong to the same cluster or not, but the answer to the query is corrupted with some probability . Let 1 > δ = 1 − 2q > 0 be the bias. In this work, we provide a polynomial time algorithm that recovers all signs correctly with high probability in the presence of noise with queries. This is the best known result for this problem for all but tiny δ, improving on the current state-of-the-art due to Mazumdar and Saha [25].","PeriodicalId":20754,"journal":{"name":"Proceedings of The Web Conference 2020","volume":"5 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2020-04-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"72992177","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Xiao Hu, Haobo Wang, Anirudh Vegesana, Somesh Dube, Kaiwen Yu, Gore Kao, Shuo-Han Chen, Yung-Hsiang Lu, G. Thiruvathukal, Ming Yin
Despite many exciting innovations in computer vision, recent studies reveal a number of risks in existing computer vision systems, suggesting results of such systems may be unfair and untrustworthy. Many of these risks can be partly attributed to the use of a training image dataset that exhibits sampling biases and thus does not accurately reflect the real visual world. Being able to detect potential sampling biases in the visual dataset prior to model development is thus essential for mitigating the fairness and trustworthy concerns in computer vision. In this paper, we propose a three-step crowdsourcing workflow to get humans into the loop for facilitating bias discovery in image datasets. Through two sets of evaluation studies, we find that the proposed workflow can effectively organize the crowd to detect sampling biases in both datasets that are artificially created with designed biases and real-world image datasets that are widely used in computer vision research and system development.
{"title":"Crowdsourcing Detection of Sampling Biases in Image Datasets","authors":"Xiao Hu, Haobo Wang, Anirudh Vegesana, Somesh Dube, Kaiwen Yu, Gore Kao, Shuo-Han Chen, Yung-Hsiang Lu, G. Thiruvathukal, Ming Yin","doi":"10.1145/3366423.3380063","DOIUrl":"https://doi.org/10.1145/3366423.3380063","url":null,"abstract":"Despite many exciting innovations in computer vision, recent studies reveal a number of risks in existing computer vision systems, suggesting results of such systems may be unfair and untrustworthy. Many of these risks can be partly attributed to the use of a training image dataset that exhibits sampling biases and thus does not accurately reflect the real visual world. Being able to detect potential sampling biases in the visual dataset prior to model development is thus essential for mitigating the fairness and trustworthy concerns in computer vision. In this paper, we propose a three-step crowdsourcing workflow to get humans into the loop for facilitating bias discovery in image datasets. Through two sets of evaluation studies, we find that the proposed workflow can effectively organize the crowd to detect sampling biases in both datasets that are artificially created with designed biases and real-world image datasets that are widely used in computer vision research and system development.","PeriodicalId":20754,"journal":{"name":"Proceedings of The Web Conference 2020","volume":"67 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2020-04-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73396072","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：