Information: Join today and make your research activities more affordable! Technical workshop participation fees and annual registration fees are available at member rates.
Notice: [Important] Announcement of Changes to Registration Fee Payment and Manuscript Upload Procedures for IEICE Technical Meetings
IEICE Technical Committee Submission System
Conference Paper's Information		 Online Proceedings
[Sign in]
Tech. Rep. Archives
 Go Top Page Go Previous   [Japanese] / [English] 

Paper Abstract and Keywords
Presentation 2026-07-14 14:15
Improving the Stability of Graph-based APT Detection with LLM Sentence Embeddings
Shota Sakamaki, Hiroki Kanezashi, Toyotaro Suzumura, Tomohiro Kudoh (UTokyo) ISEC2026-41 SITE2026-37 BioX2026-64 HWS2026-37 ICSS2026-56 EMM2026-48
Abstract (in Japanese) (See Japanese page) 
(in English) Advanced Persistent Threat (APT) attacks threaten modern information systems through stealthy，multi-stage activities that closely resemble benign behavior．Provenance-based intrusion detection systems are a promising approach for detecting such attacks by representing audit logs as graphs．However，many existing graph-based methods rely on shallow categorical features and fail to exploit the semantic information contained in raw log messages．In this paper，we propose a graph-based APT detection method that incorporates Large Language Model (LLM)-generated log representations into provenance graph learning．Our method replaces conventional categorical edge features with sentence embeddings derived from raw audit logs，allowing the model to capture fine-grained behavioral context．We evaluate the proposed method on the Unicorn Wget and DARPA TC E3 datasets．The results show that LLM-based semantic edge features improve detection performance for stealthy attacks and enhance result stability，while also revealing remaining challenges in heterogeneous environments.
Keyword (in Japanese) (See Japanese page) 
(in English) APT detection / provenance graph / graph neural network / LLM sentence embedding / anomaly detection / / /  
Reference Info. IEICE Tech. Rep., vol. 126, no. 111, ICSS2026-56, pp. 190-197, July 2026.
Paper # ICSS2026-56 
Date of Issue 2026-07-06 (ISEC, SITE, BioX, HWS, ICSS, EMM) 
ISSN Online edition: ISSN 2432-6380
Copyright
and
reproduction		 All rights are reserved and no part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Notwithstanding, instructors are permitted to photocopy isolated articles for noncommercial classroom use without fee. (License No.: 10GA0019/12GB0052/13GB0056/17GB0034/18GB0034)
Download PDF ISEC2026-41 SITE2026-37 BioX2026-64 HWS2026-37 ICSS2026-56 EMM2026-48

Conference Information
Committee BioX HWS ISEC SITE ICSS EMM IPSJ-CSEC IPSJ-SPT 
Conference Date 2026-07-13 - 2026-07-15 
Place (in Japanese) (See Japanese page) 
Place (in English) Sapporo Convention Center 
Topics (in Japanese) (See Japanese page) 
Topics (in English)  
Paper Information
Registration To ICSS 
Conference Code 2026-07-BioX-HWS-ISEC-SITE-ICSS-EMM-CSEC-SPT-SEC 
Language Japanese 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) Improving the Stability of Graph-based APT Detection with LLM Sentence Embeddings 
Sub Title (in English)  
Keyword(1) APT detection  
Keyword(2) provenance graph  
Keyword(3) graph neural network  
Keyword(4) LLM sentence embedding  
Keyword(5) anomaly detection  
Keyword(6)  
Keyword(7)  
Keyword(8)  
1st Author's Name Shota Sakamaki  
1st Author's Affiliation The University of Tokyo (UTokyo)
2nd Author's Name Hiroki Kanezashi  
2nd Author's Affiliation The University of Tokyo (UTokyo)
3rd Author's Name Toyotaro Suzumura  
3rd Author's Affiliation The University of Tokyo (UTokyo)
4th Author's Name Tomohiro Kudoh  
4th Author's Affiliation The University of Tokyo (UTokyo)
5th Author's Name  
5th Author's Affiliation ()
6th Author's Name  
6th Author's Affiliation ()
7th Author's Name  
7th Author's Affiliation ()
8th Author's Name  
8th Author's Affiliation ()
9th Author's Name  
9th Author's Affiliation ()
10th Author's Name  
10th Author's Affiliation ()
11th Author's Name  
11th Author's Affiliation ()
12th Author's Name  
12th Author's Affiliation ()
13th Author's Name  
13th Author's Affiliation ()
14th Author's Name  
14th Author's Affiliation ()
15th Author's Name  
15th Author's Affiliation ()
16th Author's Name  
16th Author's Affiliation ()
17th Author's Name  
17th Author's Affiliation ()
18th Author's Name  
18th Author's Affiliation ()
19th Author's Name  
19th Author's Affiliation ()
20th Author's Name  
20th Author's Affiliation ()
21st Author's Name  
21st Author's Affiliation ()
22nd Author's Name  
22nd Author's Affiliation ()
23rd Author's Name  
23rd Author's Affiliation ()
24th Author's Name  
24th Author's Affiliation ()
25th Author's Name  
25th Author's Affiliation ()
26th Author's Name / /
26th Author's Affiliation ()
()
27th Author's Name / /
27th Author's Affiliation ()
()
28th Author's Name / /
28th Author's Affiliation ()
()
29th Author's Name / /
29th Author's Affiliation ()
()
30th Author's Name / /
30th Author's Affiliation ()
()
31st Author's Name / /
31st Author's Affiliation ()
()
32nd Author's Name / /
32nd Author's Affiliation ()
()
33rd Author's Name / /
33rd Author's Affiliation ()
()
34th Author's Name / /
34th Author's Affiliation ()
()
35th Author's Name / /
35th Author's Affiliation ()
()
36th Author's Name / /
36th Author's Affiliation ()
()
Speaker Author-1 
Date Time 2026-07-14 14:15:00 
Presentation Time 25 minutes 
Registration for ICSS 
Paper # ISEC2026-41, SITE2026-37, BioX2026-64, HWS2026-37, ICSS2026-56, EMM2026-48 
Volume (vol) vol.126 
Number (no) no.107(ISEC), no.108(SITE), no.109(BioX), no.110(HWS), no.111(ICSS), no.112(EMM) 
Page pp.190-197 
#Pages 8 
Date of Issue 2026-07-06 (ISEC, SITE, BioX, HWS, ICSS, EMM) 

[Return to Top Page]

[Return to IEICE Web Page]
The Institute of Electronics, Information and Communication Engineers (IEICE), Japan