15:57:06 RRSAgent has joined #publishingcg 15:57:10 logging to https://www.w3.org/2026/04/23-publishingcg-irc 15:57:10 RRSAgent, make logs Public 15:57:11 Meeting: Publishing Community Group 15:58:37 chair: wolfgang 15:58:46 scribe: gautierchomel_ 15:59:17 present+ 15:59:30 present+ sebastian 15:59:35 present+ wolfgang 15:59:46 present+ nataly 16:00:14 present+ Anna 16:00:25 present+ Titusz 16:02:33 present+ laurentlm 16:03:30 present+ charles 16:04:25 wolfgang: hello, let's start to ear from Anna. 16:06:18 Anna: sharinbg the screen, 1D345 is a french company in the french Alpes, our core expereince is automatic language, we provide plagiarism detection, AI detection, content reuse, comparison 16:06:55 Anna: all data in France, GDPR compliant and confidentiality. 16:07:32 Anne: simultaneous analysis of document in depth analysis 16:08:30 Anna: we use machine learning, built our own model, find similarities to detect supsicious texts. It allows us to go deeper in analysis. 16:10:50 Anna: database of indexed contants through partnerships with publishers around the world. 16:12:26 present+ wolfgang 16:12:42 Anna: to detect AI generated contents we are following tones, context and cultural language aspects. We reach 98% reliability with less than 2 % positives. 16:13:24 Anna: paper mills markers, we detect tortured phrases and keywords to promote scientific integrity. 16:14:43 Anna: the solutions can be used at each stage, submission, peer review, to make sure we also catch revisions, and at publication. 16:15:49 Anna: incoming features, references and bibliography checking. 16:17:11 wolfgang: now let's hear from Titusz. 16:17:19 present+ Giacomo 16:17:27 present+JimS 16:18:20 Titusz: two hats for me, ISCC and Amlet. 16:19:08 Titusz: the problem is fraud patterns on online book platforms. 16:19:46 Titusz: ISO standardised ISCC open, vendor neutral and transparent. 16:20:57 ... we also develop a ISCC discovery protocol, a neutral interconnection protocol for cross registry content of authoritative metadata. 16:22:25 ... ISCC looks like a letter and numbers code generated from metadata similarity, semantic similarity, syntactic similarity, data similarity and intergity 16:23:42 ... each part of the code is individual, and ISCC combines them. 16:24:20 Titusz: demo is available at covers.iscc.io sponsored by amlet. 16:26:43 ... it indexes 3 million book covers. The initiative intent to be cross sector so it needs to be light to run. I search a title, find a cover, search for it in google, drop it to generate an ISCC. It matchs and details the match. 16:29:56 Titusz: it's an open standard so you can use it or work with a provider. Help us spread the word. 16:32:39 Titusz: Amlet on it's side is to fix the tdm reservation protocole. It uses ISCC to allow companies to identify if a title is available or not for text and data mining. It is capable to identify bokks, parts, translations, 16:33:34 Titusz: as a demo, I translate a part of that book in japanese, and it's matched! 16:33:59 present+ gregoriopellegrino 16:34:45 Titusz: AMlet is currently on waiting list, please join to get your catalog recorded. 16:34:56 wolfgang: Now Sebastian 16:35:50 Sebastian: back in 2023, we started a TF related to anti counterfeit, I try to update where we left so we can resume. 16:36:53 Sebastian: I participated in ISCC and i am the ceo of Liccium. 16:38:14 Sebastian: so in 2023 we started to receive concerns of publishing markets finding illegietiumates versions of theyre books on plateforms, both on hard copies and digital, the seller account appeared not to be the original right holder. 16:38:43 ... the problem has probably significantley increased because of the capacity of generative AI. 16:41:03 ... We identified Full republication, partial republications (original + fake press), Imitative new content presneted as associated with a brand or author, using metadata to propose false press, there is also Altered content. 16:42:40 ... we need publishers to look at this problem and involved so we can identify authoritative works and track the fraudulents ones. 16:43:17 ... in some cases, the same false ebook is proposed by several accounts with different names. 16:44:26 Sebastian: it's a challenge for publishers as they assume the cost of this identification and tracking. There is a need for standardised interoperable way to ease the path. 16:45:12 Sebastian: the original needs to be known, verifiable declarations to be attached to it. 16:46:28 Sebastian: this is why the anti counterfeit TF was set up, it has the task to document patterns, develop shared terminology, identify technical approachs, exchange best practices, explore collaborations and asses possible standardisation opportunities. 16:48:23 laurentlm: you have technologies that enable finding identical contents and metadata. One problem seems to be prevalence, sellers, plateforms, are not able to know all genuine sellers of one book. 16:49:42 Sebastian: first original publisher must claim his rights, and put pressure on retails plateforms to make them recognize this. 16:51:57 Sebastian: the technical solution exist, today problem is the will. 16:54:04 titusz: the technology surface signals, but the interpretation of this signal depend on the exhaustive available data and investement of stakeholders. Publishers needs to be together to effectively track counterfeits. 16:55:06 Sebastian: today the burden is on publishers teams to find and claim. 16:56:14 Anna: it's difficult to have pletforms investing in this because at the end they earn money from fakes as well, this is a question, how to engage them? 16:56:38 laurentlm: empty books can damage theyre reputation (for plateforms) 16:57:59 Sebastian: It happens with print books and that does not seem to be considered important by plateforms. 16:58:26 Titusz: it is easy to scrap bags of metadats from those plateforms. 16:59:16 Laurentlm: so the tech exist, the will is misisng. 17:00:14 Anna: I'm happy to see complementarities in our approachs, verifying on the publishing level and on the distribution level. Both approachs are necesary. 17:01:38 Anna: there are also means to identify writing styles in case two authors claim being orinials creators of the same text 17:06:02 rrsagent, make logs public 17:07:20 zakim, end meeting 17:07:20 As of this point the attendees have been gautierchomel_, sebastian, wolfgang, nataly, Anna, Titusz, laurentlm, charles, Giacomo, JimS, gregoriopellegrino 17:07:24 RRSAgent, please draft minutes v2 17:07:25 I have made the request to generate https://www.w3.org/2026/04/23-publishingcg-minutes.html Zakim 17:07:32 I am happy to have been of service, gautierchomel_; please remember to excuse RRSAgent. Goodbye 17:07:32 Zakim has left #publishingcg 17:07:40 rrrsagent, draft minutes 17:08:13 rrsagent, publish minutes 17:08:15 I have made the request to generate https://www.w3.org/2026/04/23-publishingcg-minutes.html wolfgang 18:06:52 gautierchomel has joined #publishingcg 19:08:58 gautierchomel has joined #publishingcg 19:25:28 gautierchomel has joined #publishingcg 19:43:13 gautierchomel has joined #publishingcg 20:07:08 gautierchomel has joined #publishingcg 20:27:54 gautierchomel has joined #publishingcg 20:45:09 gautierchomel_ has joined #publishingcg 21:02:24 gautierchomel has joined #publishingcg 21:18:34 gautierchomel has joined #publishingcg 21:34:49 gautierchomel has joined #publishingcg 21:51:04 gautierchomel has joined #publishingcg 23:08:32 gautierchomel has joined #publishingcg 23:28:23 gautierchomel has joined #publishingcg 23:44:37 gautierchomel has joined #publishingcg