US 11,657,216 B2
Input text management
Fei Fan, Xi'an (CN); Lin Yan Wu, Xi'an (CN); Xue Ying Zhang, Xi'an (CN); Jing Xu, Xi'an (CN); Yan Yan Han, Xi'an (CN); Yun Bo Zhang, Weifang (CN); and Tao Wen, Beijing (CN)
Assigned to International Business Machines Corporation, Armonk, NY (US)
Filed by International Business Machines Corporation, Armonk, NY (US)
Filed on May 20, 2021, as Appl. No. 17/325,457.
Prior Publication US 2022/0374587 A1, Nov. 24, 2022
Int. Cl. G06F 40/177 (2020.01); G06N 3/08 (2023.01)
CPC G06F 40/177 (2020.01) [G06N 3/08 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method comprising:
analyzing historical text data of a user to determine a frequency and a type of text modifications made by the user on particular key elements within the historical text data;
assigning the user to a first cluster of a plurality of clusters based on the analysis of the historical text data associated with the user, wherein the user is assigned to the first cluster based on the frequency and the type of text modifications made by the user on the particular key elements;
receiving input text from the user;
extracting a set of key elements from the input text;
referencing, a structured table mapping each of the plurality of clusters to a respective text modification scheme within the structured table; and
modifying, in response to determining that the user is assigned to the first cluster, the input text according to the referenced structured table mapping the first cluster to a first text modification scheme within the structured table, wherein the first text modification scheme includes a first mapping of the extracted key elements to text modifications.