# US 8,751,678 B2 · Method of generating a web feed and an associated system > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 8,751,678 B2 - Document ID: US8751678B2 - Status: Granted - Kind: B2 - Application number: 12/940,545 - Filed: 2010-11-05 - Published: 2011-05-05 - Granted: 2014-06-10 - Earliest claimed: 2009-11-05 - Assignees: Canon Kabushiki Kaisha - Applicants: Romain Bellessort; Youenn Fablet; Hervé Ruellan - Inventors: Romain Bellessort; Youenn Fablet; Hervé Ruellan - Web page: https://patentgrep.com/patent/US8751678B2 ## CPC - G06F 16/958: Organisation or management of web site content, e.g. publishing, maintaining pages or automatic linking - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) ## Abstract The present invention concerns a method of generating a syndication web feed (10) composed of entries (12) corresponding to content elements, and an associated system. The method comprises the steps consisting of: determining (E110, E221, E300) at least one activity parameter (A) representing a frequency of obtaining content elements to be published; calculating (E620, E770) at least one similarity value (SI) between a content element (IT) to be published in the web stream (10) and at least one entry (E) in said web stream; selecting (E130, E640) an entry (EP) in said web stream according to the at least one calculated similarity value (SI) and the at least one determined activity parameter (A); adding (E140) said content element to be published (IT) to said selected entry. ## Claims ### Claim 1 A method of generating a syndication web feed composed of entries corresponding to content elements, comprising: determining at least one activity parameter representing a quantity of new content elements to be published in the web feed that arrive in a given interval of time; calculating at least one similarity value between a new content element to be published in the web feed and at least one existing entry in said web feed; deciding whether or not to group the new content element with an existing entry in said web feed, comprising (i) selecting an existing entry in said web feed, if any, that has a calculated similarity value with the new content element greater than a similarity threshold, said similarity threshold being adjusted in accordance with the at least one determined activity parameter such that the similarity threshold increases when the at least one determined activity parameter decreases and the similarity threshold decreases when the at least one determined activity parameter increases, and (ii) making the decision with regard to the selected existing entry by determining whether or not the calculated similarity value is above or below the adjusted similarity threshold; and adding said new content element to be published to said selected existing entry of the web feed, if any, in a case where the deciding step decides to group the new content element with the selected existing entry, and otherwise adding said new content element to be published in the web feed as a new entry in a case where the deciding step decides not to group the new content element with an existing entry. ### Claim 2 (depends on 1) The method according to claim 1, wherein selecting said existing entry comprises comparing the calculated similarity value with at least one similarity threshold that depends on said activity parameter. ### Claim 3 (depends on 2) The method according to claim 2, wherein selecting the existing entry further comprises determining the entry having a greatest similarity value with the new content element to be published, the greatest similarity value being higher than at least one similarity threshold relating to the selected entry. ### Claim 4 (depends on 2) The method according to claim 2, wherein said at least one similarity threshold increases when a frequency of obtaining the content elements to be published decreases. ### Claim 5 (depends on 2) The method according to claim 2, wherein said at least one similarity value is calculated from at least one content element attribute chosen from a set comprised of a title, a web address at which the content element can be consulted, a key word, a content category and information on the origin of the content element. ### Claim 6 (depends on 1) The method according to claim 1, wherein a particular similarity threshold is associated with each entry in the web feed, and selecting the existing entry comprises comparing at least one said similarity value calculated for an entry, with said particular threshold associated with the entry. ### Claim 7 (depends on 1) The method according to claim 1, wherein said new content elements are organized by classes and at least one of said classes is associated with each entry in the web feed according to the content elements corresponding to the entry, wherein there is associated, with each class, a class similarity threshold that depends on an activity parameter relating to said class, and wherein selecting the existing entry depends on the class similarity thresholds. ### Claim 8 (depends on 7) The method according to claim 7, wherein the class similarity threshold associated with a class is updated according to a ratio between the activity parameter relating to said class and an average activity parameter relating to an average activity of all the classes. ### Claim 9 (depends on 7) The method according to claim 7, further comprising grouping entries associated with a class in the same entry when the class similarity threshold associated with the class decreases. ### Claim 10 (depends on 7) The method according to claim 7, further comprising splitting an entry associated with a class into a plurality of entries when the class similarity threshold associated with the class increases. ### Claim 11 (depends on 10) The method according to claim 10, wherein the content elements corresponding to said entry to be separated are split into a first group of content elements whose similarity with said entry is greater than or equal to a new increased similarity threshold, and a second group of content elements whose similarity of value is below the new increased similarity threshold. ### Claim 12 (depends on 7) The method according to claim 7, further comprising grouping, in the same entry of the web feed, several entries whose associated activity parameters are the lowest. ### Claim 13 (depends on 1) The method according to claim 1, wherein the content elements are organized in hierarchical categories, and said similarity value with an entry corresponds to a number of hierarchical levels in common between a new content element and at least one content element corresponding to said entry. ### Claim 14 (depends on 1) The method according to claim 1, wherein selecting the existing entry further depends on an indication of seniority of the entries in said web feed. ### Claim 15 (depends on 1) A non-transitory computer readable storage medium, able to be read by a computer system, comprising computer executable instructions for a computer program adapted to implement the method according to claim 1 when the program is loaded into and executed by the computer system. ### Claim 16 A device for generating a syndication web feed composed of entries corresponding to content elements, comprising: a determination means constructed to determine at least one activity parameter representing a quantity of new content elements to be published in the web feed that arrive in a given interval of time; a calculation means constructed to calculate at least one similarity value between a new content element to be published in the web feed and at least one existing entry of said web feed; a deciding means constructed to decide whether or not to group the new content element with an existing entry in said web feed, comprising (i) selecting an existing entry of said web feed, if any, that has a calculated similarity value with the new content element greater than a similarity threshold, said similarity threshold being adjusted in accordance with the at least one determined activity parameter such that the similarity threshold increases when the at least one determined activity parameter decreases and the similarity threshold decreases when the at least one determined activity parameter increases, and (ii) making the decision with regard to the selected existing entry by determining whether or not the calculated similarity value is above or below the adjusted similarity threshold; and an addition means constructed to add said new content element to be published to said selected existing entry of the web feed, if any, in a case where the deciding means decides to group the new content element with the selected existing entry and otherwise adding said new content element to be published in the web feed as a new entry in a case where the deciding means decides not to group the new content element with an existing entry.