Document retrieval/identification using topics

Patent metadata

Field	Value
Publication number	US-2016232157-A1
Application number	US-201514615156-A
Country	US
Kind code	A1
Filing date	Feb 5, 2015
Priority date	Feb 5, 2015
Publication date	Aug 11, 2016
Grant date	—

Abstract

A system for retrieving/identifying a document comprising text stored in a document repository is described. A memory stores a graphical structure comprising a first plurality of nodes each representing a person, and a second plurality of nodes each representing a document in the document repository, the nodes being connected by edges according to automatically observed interactions between the represented people and documents. At least some of the nodes have one or more annotations each denoting a topic. A node relatedness calculator computes distances between nodes of the graphical structure using the topic annotations. An input receives an identifier of a user who is represented by one of the first plurality of nodes. An identifier/retriever identifies one or more documents from the document repository by using the identifier and using the computed distances between nodes.

First claim

1 . A system for retrieving/identifying a document comprising text stored in a document repository comprising:
a memory storing a graphical structure comprising a first plurality of nodes each representing a person, and a second plurality of nodes each representing a document in the document repository, the nodes being connected by edges according to automatically observed interactions between the represented people and documents, at least some of the nodes having one or more annotations each denoting a topic, an interaction of the interactions at least partially based on at least one of:
a consumption activity by a person represented by a first node of the first plurality of nodes of a document represented by a first node of the second plurality of nodes, or
a relationship between a first person represented by the first node of the first plurality of nodes, and a second person represented by a second node of the first plurality of nodes;
a node relatedness calculator arranged to compute distances between nodes of the graphical structure using the topic annotations; an input arranged to receive at least an identifier of a user who is represented by one of the first plurality of nodes; and an identifier/retriever arranged to identify one or more documents from the document repository by using the identifier and using the computed distances between nodes.

2 . The system of claim 1 wherein the memory stores the graphical structure with annotations to nodes at least some of which are words which do not occur in the text of the document represented by the node.

3 . The system of claim 1 wherein the identifier/retriever is arranged to identify the one or more documents by using the identifier to locate one of the first plurality of nodes by comparing the identifier and data stored with the first plurality of nodes.

4 . The system of claim 3 wherein the node relatedness calculator is arranged to compute the distances from the located one of the first plurality of nodes to other nodes in the graphical structure.

5 . The system of claim 3 wherein the identifier/retriever is arranged to identify one or more second ones of the first plurality of nodes by using the identifier and using the computed distances between nodes.

6 . The system of claim 1 wherein the input is arranged to receive a query from end user equipment of the user, and wherein the identifier/retriever is arranged to identify the one or more documents from the document repository by using the query.

7 . The system of claim 1 comprising a topic detector arranged to automatically detect at least one topic of a document of the document repository and to save the topic as an annotation of a node of the graphical structure and/or as an annotation of the document in the document repository.

8 . The system of claim 1 comprising a topic detector arranged to automatically detect at least one topic of a document of the document repository using an index of words, where words in the index have a plurality of associated topics weighted using frequency of observed co-occurrence of the word and associated topics.

9 . The system of claim 8 comprising the index, the index having been constructed automatically without human identification of topics.

10 . The system of claim 7 comprising a user node topic aggregator arranged to aggregate or select amongst topics of documents of the document repository which have been authored or consumed by the same user, the topics of the authored or consumed documents having been automatically detected by the topic detector.

11 . The system of claim 1 comprising an updater arranged to update the graphical structure by adding, editing or deleting nodes and/or edges of the graph according to automatically observed document authoring and document consumption events.

12 . The system of claim 11 wherein the updater is arranged to update the graphical structure by adding, editing or deleting the annotations.

13 . A computer-implemented method of retrieving/identifying a document comprising text stored in a document repository comprising:
storing, at a memory, a graphical structure comprising a first plurality of nodes each representing a person, and a second plurality of nodes each representing a document in the document repository, the nodes being connected by edges according to automatically observed interactions between the represented people and documents, at least some of the nodes having one or more annotations each denoting a topic, an interaction of the interactions at least partially based on at least one of:
a consumption activity by a person represented by a first node of the first plurality of nodes of a document represented by a first node of the second plurality of nodes, or
a relationship between a first person represented by the first node of the first plurality of nodes, and a second person represented by a second node of the first plurality of nodes;
computing, at a processor, distances between nodes of the graphical structure using the topic annotations; receiving an identifier of a user who is represented by one of the first plurality of nodes; and automatically identifying one or more documents from the document repository by using the identifier and using the computed distances between nodes.

14 . The method of claim 13 comprising storing the graphical structure with annotations to nodes at least some of which are words which do not occur in the text of the document represented by the node.

15 . The method of claim 13 comprising automatically detecting at least one topic of a document of the document repository using an index of words, where words in the index have a plurality of associated topics weighted using frequency of observed co-occurrence of the word and associated topics in a corpus of documents which is the same or different from the document repository.

16 . The method of claim 15 comprising updating the index by observing co-occurrence of words and associated topics in new documents.

17 . One or more computer storage media with device-executable instructions that, when executed by a computing system, direct the computing system to perform steps comprising:
storing, at a memory, a graphical structure comprising a first plurality of nodes each representing a person, and a second plurality of nodes each representing a document in the document repository, the nodes being connected by edges according to automatically observed interactions between the represented people and documents, at least some of the nodes having one or more annotations each denoting a topic, an interaction of the interactions at least partially based on at least one of:
a consumption activity by a person represented by a first node of the first plurality of nodes of a document represented by a first node of the second plurality of nodes, or
a relationship between a first person represented by the first node of the first plurality of nodes, and a second person represented by a second node of the first plurality of nodes;
computing, at a processor, distances between nodes of the graphical structure using the topic annotations; receiving an identifier of a user who is represented by one of the first plurality of nodes; and automatically identifying one or more documents from the document repository, or people from the graphical structure, by using the identifier and using the computed distances between nodes.

18 . The one or more computer storage media of claim 17 with device-executable instructions that, when executed by a computing system, dire

Assignees

• Microsoft Technology Licensing Llc

Inventors

• Mansour Riham Hassan Abdel-Moneim
• Ashour Ahmed Adel Mohamed Abdel Kader
• Abdelwahab El Baz Hesham Saad Mohamed

Classifications

• G06F16/3334

  Selection or weighting of terms from queries, including natural language queries · CPC title

• G06F16/31 Primary

  Indexing; Data structures therefor; Storage structures · CPC title

• G06F18/22

  Matching criteria, e.g. proximity measures · CPC title

• G06K9/00483

  Physics · mapped topic

• G06K9/6215

  Physics · mapped topic