{"title":"ASSIST:大型时间图中重要结构变化的自动总结","authors":"C. Chelmis, R. Dani","doi":"10.1145/3091478.3091518","DOIUrl":null,"url":null,"abstract":"Detecting outliers and anomalies in data is vital in numerous applications in areas such as security, finance, health care, and online social media. Such dynamic systems can be modeled as graphs that change over time. Even though considerable work has been performed on finding points in time at which a network notably differs from its past, little work has been done on characterizing or explaining such changes. However, in the era of big data where networked data are getting bigger and bigger, being able to summarize such changes is key for sensemaking and root cause analysis. To address this gap, we present a novel approach to summarize significant structural changes in large temporal graphs. Specifically, we propose an efficient approach to help the user understand sharp changes in the structure of the network by presenting to her only a summary of key subgraphs that contribute most to the change. Extensive evaluation on real-world datasets with ground truth demonstrates both quantitatively and qualitatively the ability of our approach to accurately detect changes and discover \"important\" structures to succinctly describe the change.","PeriodicalId":165747,"journal":{"name":"Proceedings of the 2017 ACM on Web Science Conference","volume":"20 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2017-06-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"2","resultStr":"{\"title\":\"ASSIST: Automatic Summarization of Significant Structural Changes in Large Temporal Graphs\",\"authors\":\"C. Chelmis, R. Dani\",\"doi\":\"10.1145/3091478.3091518\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Detecting outliers and anomalies in data is vital in numerous applications in areas such as security, finance, health care, and online social media. Such dynamic systems can be modeled as graphs that change over time. Even though considerable work has been performed on finding points in time at which a network notably differs from its past, little work has been done on characterizing or explaining such changes. However, in the era of big data where networked data are getting bigger and bigger, being able to summarize such changes is key for sensemaking and root cause analysis. To address this gap, we present a novel approach to summarize significant structural changes in large temporal graphs. Specifically, we propose an efficient approach to help the user understand sharp changes in the structure of the network by presenting to her only a summary of key subgraphs that contribute most to the change. Extensive evaluation on real-world datasets with ground truth demonstrates both quantitatively and qualitatively the ability of our approach to accurately detect changes and discover \\\"important\\\" structures to succinctly describe the change.\",\"PeriodicalId\":165747,\"journal\":{\"name\":\"Proceedings of the 2017 ACM on Web Science Conference\",\"volume\":\"20 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2017-06-25\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"2\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Proceedings of the 2017 ACM on Web Science Conference\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1145/3091478.3091518\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings of the 2017 ACM on Web Science Conference","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1145/3091478.3091518","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
ASSIST: Automatic Summarization of Significant Structural Changes in Large Temporal Graphs
Detecting outliers and anomalies in data is vital in numerous applications in areas such as security, finance, health care, and online social media. Such dynamic systems can be modeled as graphs that change over time. Even though considerable work has been performed on finding points in time at which a network notably differs from its past, little work has been done on characterizing or explaining such changes. However, in the era of big data where networked data are getting bigger and bigger, being able to summarize such changes is key for sensemaking and root cause analysis. To address this gap, we present a novel approach to summarize significant structural changes in large temporal graphs. Specifically, we propose an efficient approach to help the user understand sharp changes in the structure of the network by presenting to her only a summary of key subgraphs that contribute most to the change. Extensive evaluation on real-world datasets with ground truth demonstrates both quantitatively and qualitatively the ability of our approach to accurately detect changes and discover "important" structures to succinctly describe the change.