MARC보기
LDR02849nmm uu200445 4500
001000000332218
00520240805170413
008181129s2018 |||||||||||||||||c||eng d
020 ▼a 9780438283725
035 ▼a (MiAaPQ)AAI10969856
040 ▼a MiAaPQ ▼c MiAaPQ ▼d 248032
0820 ▼a 004
1001 ▼a Moon, Changsung.
24510 ▼a Predictive Modeling of Complex Graphs as Context and Semantics Preserving Vector Spaces.
260 ▼a [S.l.] : ▼b North Carolina State University., ▼c 2018
260 1 ▼a Ann Arbor : ▼b ProQuest Dissertations & Theses, ▼c 2018
300 ▼a 93 p.
500 ▼a Source: Dissertation Abstracts International, Volume: 79-12(E), Section: B.
500 ▼a Adviser: Nagiza F. Samatova.
5021 ▼a Thesis (Ph.D.)--North Carolina State University, 2018.
520 ▼a Predictive modeling of complex graphs is a process that uses complex graph data and probability theory to forecast outcomes. Relational data, such as social networks and knowledge bases, can be represented as complex graphs with directed, multi-
520 ▼a This work initially addresses two research challenges with regards to predictive analysis in complex graphs: 1) automatic completion of user-intended actions and 2) automatic knowledge graph (KG) completion, i.e., inference of missing entities a
520 ▼a Typically, frequency analysis based models andMarkov models have been applied to address the task of predicting user actions. Although the existing models have been shown to have a reasonable predictive strength, they have issues capturing seman
520 ▼a In this dissertation, we seek to improve both prediction of user actions and KG completion by inventing vector space embedding methods. We first propose an online method, Frequency Vector (FVEC) prediction, that predicts next actions by combinin
520 ▼a For the prediction of missing entities and relation types in KGs, we present a contextual embedding method CONTE that learns the vector embeddings of entities and relation types while taking contextual relation types into account. Contextual re
520 ▼a Finally, we present an embedding method inferring missing entity types, which refer to collections of entities that share common definitions (e.g., /music/artist). In addition to inference of missing entities and relation types, inferring missi
590 ▼a School code: 0155.
650 4 ▼a Computer science.
650 4 ▼a Artificial intelligence.
690 ▼a 0984
690 ▼a 0800
71020 ▼a North Carolina State University. ▼b Computer Science.
7730 ▼t Dissertation Abstracts International ▼g 79-12B(E).
773 ▼t Dissertation Abstract International
790 ▼a 0155
791 ▼a Ph.D.
792 ▼a 2018
793 ▼a English
85640 ▼u http://www.riss.kr/pdu/ddodLink.do?id=T15001284 ▼n KERIS
980 ▼a 201812 ▼f 2019
990 ▼a 관리자