T 0955/20 - (II) Better search results are not technical

Key points

• The appeal is against the refusal of a patent application with applicant Google.
• The Board, on inventive step: " Furthermore, even if distinguishing features (b) and (c) could be seen to plausibly improve the "quality" of the translated query in the sense that it leads to search results that are "better" in the sense of closer in meaning to the desired search results, this would not be a technical effect (see decisions T 1569/05, Reasons 3.4 to 3.7; T 598/14, Reasons 2.3 and 2.4)." 
• T 1569/05: " The appellants regard the invention as belonging to the technical field of utilizing a natural language as a search input. In their view, the invention makes a technical contribution in a field not excluded from patentability," " In the Board's view, neither the mathematical model of meaning according to D3 nor the modified model according to the invention are within the technical area, since only the meaning of the words determines how they are represented, stored and selected, and since mathematical algorithms completely define the processing. In this respect the present invention is similar to the case T 52/85 - Listing of semantically related expressions/IBM (not published in OJ EPO), where the deciding board held that automatically generating a list of expressions semantically related to an input linguistic expression is basically not of a technical nature but a matter of the meaning of those expressions, ie of their abstract linguistic information content." 
•  T 52/85: "Such a semantical relationship is basically not of a technical nature but a matter of the meaning of those expressions, i.e. of their abstract linguistic information content; it does not relate to any physical entity. A semantical relationship can be found by performing mental acts only, with no technical means involved. This does not necessarily mean that any system automatically concatenating, in place of a human being, semantically related expressions to a list is excluded from patentability. Rather, this will depend on whether the manner in which it is automated, involves features which make a contribution in a field outside the range of matters excluded from patentability under Article 52(2) and (3) EPC."
• "the appellant's observation on document D1 does not call into question the board's finding that the distinguishing features provide no technical contribution over document D1. Hence, the subject-matter of claim 1 lacks inventive step (Article 56 EPC)."

EPO T 0955/20 

The link to the decision is provided after the jump, as well as (an extract of) the text of the decision.

https://www.epo.org/law-practice/case-law-appeals/recent/t200955eu1.html

4. The invention as defined by claim 1

4.1 The application generally relates to performing searches using queries that are written in a character set or language different from the character set or language of the documents being searched by first translating the queries into the appropriate character set or language.

4.2 Claim 1 is directed to such a search method, with the restriction that the character sets are different and "an actual language of the query is not changed".

First, a query, written in a first format, is received from a user and translated into a second format. The document database is then searched using the translated query, and search results written in the second format are returned to the user. The user may then select a search result.

4.3 The translation is performed by means of a "probabilistic dictionary" which maps terms in the first format to terms in the second format. The first and second formats refer to the character sets in which the terms are expressed, for example romaji and kanji (see page 13, line 27, to page 14, line 30, of the description).

4.4 The search-result selections obtained from the user are used to modify at least one mapping probability.

5. Inventive step

5.1 Document D1 relates to searching a database which contains documents written in different languages (see abstract and paragraph [0002]). When a user enters a query with keywords in a first language, a keyword dictionary is used to translate the query into a query with keywords in a second language (paragraph [0026]; Figures 3 and 5), which may involve translating into a different character set (paragraph [0028]). Each query is then applied to an inverted index corresponding to its respective language to generate search results (ibid.). The user may then select search results, which implies that the search results are shown to the user (ibid.).

5.2 The subject-matter of claim 1 differs from the disclosure of document D1 in that

(a) the first and second formats are not different languages but different character sets for the same language;

(b) the dictionary is a "probabilistic" dictionary;

(c) search-results selections are used to modify at least one mapping probability of the dictionary.

5.3 Distinguishing feature (a) expresses the non-technical, linguistic fact that the same language, such as Japanese, may have different representations in different character sets, such as romaji and kanji. It merely concerns the cognitive content of the dictionary (mappings from romaji terms to kanji terms instead of, for example, Greek terms to English terms), not its technical implementation. It therefore cannot contribute to an inventive step.

5.4 Distinguishing features (b) and (c) express that the mappings in the dictionary have probabilities assigned to them, which are updated in response to selections of search results made by the user.

Claim 1 does not specify that the probabilities are used, let alone how they are used. For this reason alone, features (b) and (c) do not contribute to any technical effect.

Furthermore, even if distinguishing features (b) and (c) could be seen to plausibly improve the "quality" of the translated query in the sense that it leads to search results that are "better" in the sense of closer in meaning to the desired search results, this would not be a technical effect (see decisions T 1569/05, Reasons 3.4 to 3.7; T 598/14, Reasons 2.3 and 2.4).

5.5 The appellant argued that features (b) and (c) contributed to solving the technical problem of "how to provide improvements in query input using data input devices that use a format/character set different from the format/character set of at least some of the documents to be searched". The Guidelines for Examination, G-II, 3.7.1, clarified that features enabling user input made a technical contribution.

However, to the extent that claim 1 solves this problem, it is also solved, in the same way, by the method disclosed in document D1, which allows the searching of documents written in a (language and) character set different from the (language and) character set in which the query was formulated by translating the query. The solution to the problem does not rely on distinguishing features (b) and (c).

5.6 At the oral proceedings before the board, the appellant insisted that document D1 was not concerned with allowing a user to input a search query with an input device that did not support the character set in which the documents being searched were written.

It is not fully clear to the board what it would mean for an input device to not support a particular character set, given that the description, on page 1, line 31, to page 2, line 14, explains that conventional methods allow the use of a standard telephone keypad to enter a query for "ben smith", i.e. a query expressed in the Roman alphabet. In any event, claim 1 is not restricted to queries that are entered by means of any particular input device. In fact, the scope of claim 1 encompasses not only methods which translate a query written in romaji (a phonetic, Roman-alphabet representation of Japanese) into kanji, i.e. from a character set directly supported by ASCII keyboards into a character set which arguably is not directly supported by such keyboards, but also methods which translate a query written in kanji into romaji.

Hence, the appellant's observation on document D1 does not call into question the board's finding that the distinguishing features provide no technical contribution over document D1.

5.7 Hence, the subject-matter of claim 1 lacks inventive step (Article 56 EPC).