公开(公告)号：US10740406B2

公开(公告)日：2020-08-11

申请号：US15100918

申请日：2013-12-06

Applicant: Hewlett-Packard Development Company, L.P.

Inventor： Georgia Koutrika ,  Dimitra Papadimitriou ,  Steven J Simske

IPC: G06F16/93 ,  G06F16/2457 ,  G06F40/20 ,  G06F40/205

Abstract: Matching of an input document to documents in a document collection is described herein. In an example, a similarity correspondence between an input document and one or more documents in a base document collection is established. A set of base document segments and a set of message types associated to document segments in the set of base document segments is provided. The set of base document segments is derived from documents in the base document collection. The input document is segmented into input document segments corresponding to message types. Segment similarity between input document segments and base document segments corresponding to the same message types is computed. The similarity correspondence between the input document and at least one document in the base document collection is based on the computed segment similarity.

公开(公告)号：US20160299891A1

公开(公告)日：2016-10-13

申请号：US15100918

申请日：2013-12-06

Applicant: HEWLETT-PACKARD DEVELOPMENT COMPANY, L.P.

Inventor： Georgia Koutrika ,  Dimitra Papadimitriou ,  Steven J Simske

IPC: G06F17/30 ,  G06F17/27

Abstract: Matching of an input document to documents in a document collection is described herein. In an example, a similarity correspondence between an input document and one or more documents in a base document collection is established. A set of base document segments and a set of message types associated to document segments in the set of base document segments is provided. The set of base document segments is derived from documents in the base document collection. The input document is segmented into input document segments corresponding to message types. Segment similarity between input document segments and base document segments corresponding to the same message types is computed. The similarity correspondence between the input document and at least one document in the base document collection is based on the computed segment similarity.

Abstract translation: 这里描述了输入文档与文档集合中的文档的匹配。 在一个示例中，建立输入文档和基本文档集合中的一个或多个文档之间的相似性对应关系。 提供了一组基本文档段和一组与基本文档段中的文档段相关联的消息类型。 基本文档段的集合是从基础文档集合中的文档导出的。 输入文档被分割成与消息类型对应的输入文档段。 计算对应于相同消息类型的输入文档段和基本文档段之间的段相似性。 输入文档与基本文档集合中的至少一个文档之间的相似性对应关系基于所计算的片段相似度。