Computing machine and template management method

A computing machine managing a plurality of templates, each defining a position, on a document, of at least one attribute contained in the document, the computing machine comprising:

• at least one processor;
• at least one storage storing thereon a plurality of programs that when executed by the at least one processor cause the computing machine to perform the steps of:
• obtaining information of a target document that has a document format, which has been subjected to a recognition process, and wherein the recognition process acquires attributes on the target document;
• calculating an optical character recognition (OCR) score and a position score of a specific attribute from the acquired attributes by comparing the information of the target document with a previously stored template, the OCR score corresponding to a degree of similarity between an acquired character string of the specific attribute and a character string of a stored attribute on a stored template and the position score corresponding to a deviation between a position of the specific attribute and a position of the stored attribute;
• determining whether a cluster of a plurality of clusters, that are predefined and each cluster corresponding to at least one other document having the document format, to which the target document can belong is present on the basis of the position score of the specific attribute of the target document by perform ing the steps of:
• calculating a distance between the position score of the target document and a center of gravity of each cluster of the plurality of clusters;
• and
• determining whether there is at least one cluster to which the target document can belong is present on the basis of a result of comparison between the distance and a threshold;
• upon determining that there is no cluster present to which the target document can belong on the basis of the result of the comparison, output an instruction to generate a new template and display a template generation screen having a selectable list of documents and a list of attributes that comprise the acquired attributes and one or more new attributes that are added upon operator input and wherein positions of the attributes listed are displayed according to document image data of a selected document for the selectable list of documents provided on the template generation screen and registering the new template based on the target document upon receiving an instruction to register the new template;
• generating a new cluster upon determining that there is no cluster present to which the target document can belong;
• and
• managing the document format of the target document, the new template and the new cluster in such a manner that the document format of the target document, the new template of the target document, and the new cluster correspond to one another.