Talk:Cloem

This page should not be speedy deleted because...

This page should not be speedily deleted because it's not a promotional article but an informative article about a new invention that can reform the whole patent policy all around the world - it's not just a product or a page for profit but a new invention to help other inventors. --Samira RZ (talk) 13:13, 24 November 2015 (UTC)

• The article is written as an advertisement with supported advertisement links.
  

Maybe the company is worthy, but the article, so far, is not. MarkYabloko 13:30, 24 November 2015 (UTC)

@MarkYabloko: I don't think it's written like an ad, it's pretty objective point of view aknowledging that this invention might reform some real big things in the subject of patenting and law. Samira RZ (talk) 13:18, 25 November 2015 (UTC)

References moved from article

I was invited here by user:Edcolins to help clean up the page a bit. For the record, I am registered US patent agent and a 10 year veteran of Wikipedia. I understand the creator of the page is new to Wikipedia, so I thought I would take it in stages. I will start by moving references here that are not reliable secondary sources and see if there are any concerns. Also, for the record, it looks to me like Cloem has an interesting concept.--Nowa (talk) 23:05, 8 December 2015 (UTC)

• Self published and anonymous "Algorithmic-patenting" Slashdot
• Self published by marketing manager Defensive "Cloems" - Cullens - Patent and Trademark Attorneys (although interesting comment about crowdsourcing in article footnote)
• Rehash of existing reference (Venture Beat): Spicy IP also doesn't support section content.
• Self published by not a patent attorney |title=Computer Generated Patent Claims: Good Thing or Bad Thing? |publisher=Mark Smith Attorneys

Unsupported material

OK, the next step is to remove unsupported material. I'll put it here. If references are found to support it, we can move it back. I usually like to start from the bottom and work my way up.--Nowa (talk) 15:48, 9 December 2015 (UTC)

Criticism

Existing reactions are diverse. While Cloem can be interpreted to be an attack of the patent system (brutalizating the permutation space), it can also appear to be part of the "Legal Tech" movement (disruptive application of new technologies to legal practices, that is to follow patent laws as Copyleft follows copyright laws). At least, Cloem's approach raises new questions and challenges for the patent systems in the world.

Prior Art

As described by the company's scientists, each created text is mathematically timestamped (trusted timestamping and/or trustless timestamping with a crypto-ledger like a blockchain).

Technology

Dictionaries are derived from Wordnet, Wikipedia, claim construction dictionariesand the analysis of 70 million patent documents.

As described by the company's scientists, Cloem’s claim general drafting model is said to follow industry "best practices" (e.g. USPTO M.P.E.P., EPO Examination Guidelines, Case Law, etc.) and to optionally implement a proprietary implementation of TRIZ (injection of invention patterns, as identified in the general patent corpus).Cloem technology page (untrusted connection)

Cloem’s algorithms also include the injection of the definition of terms.

Grammatical algorithms include the combinatorial permutation of clauses of a patent claim, the handling of time (e.g. causality of events), the transcoding of claim categories (e.g. method claim into system claim and vice-versa) and the creation of intermediate generalizations (by inserting combinations of hyponyms and hyperonyms).

Untrusted connection to Cloem web site

When I tried to connect to Cloem's web site via Firefox or Safari, I got an "untrusted" connection warning. For the safety of other Wikipedia readers, I am moving the connections here until they get a trusted connection.--Nowa (talk) 16:10, 9 December 2015 (UTC)

External links

• Official website

I have the same problem. --Edcolins (talk) 20:24, 9 December 2015 (UTC)

Original Research

The Epistemology section looks the the good start of a paper for publication. Once it's published, it might be a good reference for this article. Until then, I'm moving it here per wp:or. Also, for the record, I happen to be a guest writer for IP Watchdog. If the original poster wants to publish said paper on IP Watchdog, I would be happy to make an introduction to the editor.--Nowa (talk) 16:39, 9 December 2015 (UTC)

Epistemology

A patent claim in theory shall be as clear and concise as possible. But as a sentence expressed in natural language, it is by nature irreducibly ambiguous. The ambiguity can be lexical, syntactical and/or semantical. By injecting definitions of words, or replacing words by their synonyms, the meaning of a patent claim can be disambiguated to some extent. The scope of protection can thereby be clarified (the boundaries of protection for private interests can be better identified by the general public, see Legal security or Rule of Law.

By injecting hyponyms or hyperonyms, the semantic meaning of an initial patent claim can be modified. Amplifying human creativity, machine can "exhaust" ideas. If a "technical pattern" emerges in the linguistic corpus (for example a new combination of words like "touch screen"), said pattern can be propagated to the entire patent linguistic corpus (i.e. it can be part of the prior art). Compared to word processing software, the proposed algorithmic patenting essentially consists in pushing thesaurus and dictionaries into patent claims. When an inventor provides an invention, he "implicitly" discloses a "class" of inventions and it is rather fair to let the machine "fill-in" the gaps.

The more inventions are patentable, the more fees can be collected by the patent system (by construction). Patent classification systems influence patentability of inventions, since an invention can be considered inventive in a specific technical domain and not in another due to the patent classification structure. By contrast, the systematic transposition of technical patterns (by analogy reasoning or by text manipulation) or the mixing-up patent of classification classes would substantially decrease the number of allowable patents.

For example, a "business method" patent claim is a "software patent" claim with a few words being changed (e.g. "bank" replaced by "server" and "order" replaced by "message"). In theory, software patents can be changed into business method patents by mere textual manipulation. More generally, by considering a parsed patent claim or sentence as a "graph" of words and by manipulating its words, the frontiers between mechanical, chemical, business, software and other types of patents are significantly blurred.

@Nowa: thank you so much! I didn't even see this. So you think I shoult put this part as its own new article? And what exactly would be good for the page you talkd about? the IP Watchdop I mean. Thanks a lot in advance for answering! Samira RZ (talk) 13:38, 12 January 2016 (UTC)

References

Issues addressed?

I believe I've addressed the issues in the "multiple issues" warning. If any other editors agree, feel free to remove the warning.--Nowa (talk) 16:51, 9 December 2015 (UTC)

I've removed the "multiple issues" warning. --Edcolins (talk) 21:43, 9 December 2015 (UTC)