Event prediction aims to predict the most possible following event given a chain of closely related context events.Previous methods based on event pairs or the entire event chain may ignore much structural and semanti...Event prediction aims to predict the most possible following event given a chain of closely related context events.Previous methods based on event pairs or the entire event chain may ignore much structural and semantic information.Current datasets for event prediction,naturally,can be used for supervised learning.Event chains are either from document-level procedural action flow,or from news sequences under the same column.This paper leverages graph structure knowledge of event triggers and event segment information for event prediction with general news corpus,and adopts the standard multiple choice narrative cloze task evaluation.The topic model is utilized to extract event chains from the news corpus to deal with training data bottleneck.Based on trigger-guided structural relations in the event chains,we construct trigger evolution graph,and trigger representations are learned through graph convolutional neural network and the novel neighbor selection strategy.Then there are features of two levels for each event,namely,text level semantic feature and trigger level structural feature.We design the attention mechanism to learn the features of event segments derived in term of event major subjects,and integrate relevance between event segments and the candidate event.The most possible next event is picked by the relevance.Experimental results on the real-world news corpus verify the effectiveness of the proposed model.展开更多
The study of dynamic networks in computer science has become crucial, given their ever-evolving nature within digital ecosystems. These networks serve as fundamental models for various networked systems, usually chara...The study of dynamic networks in computer science has become crucial, given their ever-evolving nature within digital ecosystems. These networks serve as fundamental models for various networked systems, usually characterized by modular structures. Understanding these structures, also known as communities, and the mechanisms driving their evolution is vital, as changes in one module can impact the entire network. Traditional static network analysis falls short of capturing the full complexity of dynamic networks, prompting a shift toward understanding the underlying mechanisms driving their evolution. Graph Evolution Rules (GERs) have emerged as a promising approach, explaining how subgraphs transform into new configurations. In this paper, we comprehensively explore GERs in dynamic networks from diverse systems with a focus on the rules characterizing the formation and evolution of their modular structures, using EvoMine for GER extraction and the Leiden algorithm for community detection. We characterize network and module evolution through GER profiles, enabling cross-system comparisons. By combining GERs and network communities, we decompose network evolution into regions to uncover insights into global and mesoscopic network evolution patterns. From a mesoscopic standpoint, the evolution patterns characterizing communities emphasize a non-homogeneous nature, with each community, or groups of them, displaying specific evolution patterns, while other networks’ communities follow more uniform evolution patterns. Additionally, closely interconnected sets of communities tend to evolve similarly. Our findings offer valuable insights into the intricate mechanisms governing the growth and development of dynamic networks and their communities, shedding light on the interplay between modular structures and evolving network dynamics.展开更多
The Internet presents numerous sources of useful information nowadays. However, these resources are drowning under the dynamic Web, so accurate finding user-specific information is very difficult. In this paper we dis...The Internet presents numerous sources of useful information nowadays. However, these resources are drowning under the dynamic Web, so accurate finding user-specific information is very difficult. In this paper we discuss a Semantic Graph Web Search (SGWS) algorithm in topic-specific resource discovery on the Web. This method combines the use of hyperlinks, characteristics of Web graph and semantic term weights. We implement the algorithm to find Chinese medical information from the Internet. Our study showed that it has better precision than traditional IR (Information Retrieval) methods and traditional search engines. Key words HITS - evolution web graph - power law distribution - context analysis CLC number TP 391 - TP 393 Foundation item: Supported by the National High-Performance Computation Fund (00303)Biography: Ye Wei-guo (1970-), male, Ph. D candidate, research direction: Web information mining, network security, artificial intelligence.展开更多
基金supported by the National Natural Science Foundation of China(71731002,71971190).
文摘Event prediction aims to predict the most possible following event given a chain of closely related context events.Previous methods based on event pairs or the entire event chain may ignore much structural and semantic information.Current datasets for event prediction,naturally,can be used for supervised learning.Event chains are either from document-level procedural action flow,or from news sequences under the same column.This paper leverages graph structure knowledge of event triggers and event segment information for event prediction with general news corpus,and adopts the standard multiple choice narrative cloze task evaluation.The topic model is utilized to extract event chains from the news corpus to deal with training data bottleneck.Based on trigger-guided structural relations in the event chains,we construct trigger evolution graph,and trigger representations are learned through graph convolutional neural network and the novel neighbor selection strategy.Then there are features of two levels for each event,namely,text level semantic feature and trigger level structural feature.We design the attention mechanism to learn the features of event segments derived in term of event major subjects,and integrate relevance between event segments and the candidate event.The most possible next event is picked by the relevance.Experimental results on the real-world news corpus verify the effectiveness of the proposed model.
基金supported by the Italian Ministry of University and Research(MUR)and the European Union–NextGenerationEU in the framework of the PRIN 2022 project“AWESOME:Analysis framework for WEb3 SOcial MEdia”–CUP:I53D23003680006.
文摘The study of dynamic networks in computer science has become crucial, given their ever-evolving nature within digital ecosystems. These networks serve as fundamental models for various networked systems, usually characterized by modular structures. Understanding these structures, also known as communities, and the mechanisms driving their evolution is vital, as changes in one module can impact the entire network. Traditional static network analysis falls short of capturing the full complexity of dynamic networks, prompting a shift toward understanding the underlying mechanisms driving their evolution. Graph Evolution Rules (GERs) have emerged as a promising approach, explaining how subgraphs transform into new configurations. In this paper, we comprehensively explore GERs in dynamic networks from diverse systems with a focus on the rules characterizing the formation and evolution of their modular structures, using EvoMine for GER extraction and the Leiden algorithm for community detection. We characterize network and module evolution through GER profiles, enabling cross-system comparisons. By combining GERs and network communities, we decompose network evolution into regions to uncover insights into global and mesoscopic network evolution patterns. From a mesoscopic standpoint, the evolution patterns characterizing communities emphasize a non-homogeneous nature, with each community, or groups of them, displaying specific evolution patterns, while other networks’ communities follow more uniform evolution patterns. Additionally, closely interconnected sets of communities tend to evolve similarly. Our findings offer valuable insights into the intricate mechanisms governing the growth and development of dynamic networks and their communities, shedding light on the interplay between modular structures and evolving network dynamics.
文摘The Internet presents numerous sources of useful information nowadays. However, these resources are drowning under the dynamic Web, so accurate finding user-specific information is very difficult. In this paper we discuss a Semantic Graph Web Search (SGWS) algorithm in topic-specific resource discovery on the Web. This method combines the use of hyperlinks, characteristics of Web graph and semantic term weights. We implement the algorithm to find Chinese medical information from the Internet. Our study showed that it has better precision than traditional IR (Information Retrieval) methods and traditional search engines. Key words HITS - evolution web graph - power law distribution - context analysis CLC number TP 391 - TP 393 Foundation item: Supported by the National High-Performance Computation Fund (00303)Biography: Ye Wei-guo (1970-), male, Ph. D candidate, research direction: Web information mining, network security, artificial intelligence.
