# US 2014/0156571 A1 · Topic models > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2014/0156571 A1 - Document ID: US20140156571A1 - Status: Published application - Kind: A1 - Application number: 14/172,771 - Filed: 2014-02-04 - Published: 2014-06-05 - Earliest claimed: 2010-10-26 - Assignees: Microsoft Corporation - Applicants: Microsoft Corporation - Inventors: Philipp Hennig; David Stern; Thore Graepel; Ralf Herbrich - Web page: https://patentgrep.com/patent/US20140156571A1 ## CPC - G06N 20/00: Machine learning - G06N 7/01: Probabilistic graphical models, e.g. probabilistic networks - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS ## Abstract Machine learning techniques may be used to train computing devices to understand a variety of documents (e.g., text files, web pages, articles, spreadsheets, etc.). Machine learning techniques may be used to address the issue that computing devices may lack the human intellect used to understand such documents, such as their semantic meaning. Accordingly, a topic model may be trained by sequentially processing documents and/or their features (e.g., document author, geographical location of author, creation date, social network information of author, and/or document metadata). Additionally, as provided herein, the topic model may be used to predict probabilities that words, features, documents, and/or document corpora, for example, are indicative of particular topics. ## Claims ### Claim 1 A method for predicting a topic of a document, comprising: processing a document representation of a document and features of the document using a topic model, the processing comprising: determining a feature/topic prediction for a feature of the document, the feature/topic prediction specifying a probability of the feature being indicative of a first topic; determining a word/topic prediction for a word within the document, the word/topic prediction specifying a probability of the word being indicative of a second topic; and determining a document/topic prediction for the document based upon the feature/topic prediction and the word/topic prediction. ### Claim 2 (depends on 1) The method of claim 1, the document/topic prediction specifying a probability of the document being indicative of a third topic. ### Claim 3 (depends on 1) The method of claim 1, the features of the document comprising author designation. ### Claim 4 (depends on 1) The method of claim 1, the features of the document comprising geographical location. ### Claim 5 (depends on 1) The method of claim 1, the features of the document comprising at least one of creation date designation or creation time designation. ### Claim 6 (depends on 1) The method of claim 1, the features of the document comprising at least one of document metadata or source of the document. ### Claim 7 (depends on 1) The method of claim 1, the features of the document comprising document length. ### Claim 8 (depends on 1) The method of claim 1, the features of the document comprising social network membership. ### Claim 9 (depends on 1) The method of claim 1, the features of the document comprising previous search queries. ### Claim 10 (depends on 1) The method of claim 1, the features of the document comprising reader designation. ### Claim 11 (depends on 1) The method of claim 1, the features of the document comprising web browsing history. ### Claim 12 (depends on 1) The method of claim 1, the features of the document comprising document type. ### Claim 13 A system for training a topic model, comprising: one or more processing units; and memory comprising instructions that when executed by at least one of the one or more processing units, perform a method comprising: for a document within a document corpus: receiving a document representation of the document and features of the document, the document representation comprising a frequency of word occurrences within the document; processing the document representation and the features using a topic model, the processing comprising: specifying a feature/topic parameter for a feature of the document, the feature/topic parameter specifying a probability of the feature being indicative of a first topic, the feature/topic parameter based upon a first uncertainty measure; and specifying a document/word/topic parameter for a word within the document, the document/word/topic parameter specifying a probability of the word being indicative of a second topic, the document/word/topic parameter based upon a second uncertainty measure; and training the topic model based upon the feature/topic parameter and the document/word/topic parameter. ### Claim 14 (depends on 13) The system of claim 13, the features of the document comprising at least one of author designation or reader designation. ### Claim 15 (depends on 13) The system of claim 13, the features of the document comprising geographical location. ### Claim 16 (depends on 13) The system of claim 13, the features of the document comprising at least one of creation date designation or creation time designation. ### Claim 17 (depends on 13) The system of claim 13, the features of the document comprising at least one of document metadata or source of the document. ### Claim 18 (depends on 13) The system of claim 13, the features of the document comprising at least one of document length or document type. ### Claim 19 (depends on 13) The system of claim 13, the features of the document comprising at least one of social network membership or web browsing history. ### Claim 20 A computer readable medium comprising instructions that when executed, perform a method for predicting a topic of a document, the method comprising: processing a document representation of a document and features of the document using a topic model, the processing comprising: determining a feature/topic prediction for a feature of the document, the feature/topic prediction specifying a probability of the feature being indicative of a first topic; determining a word/topic prediction for a word within the document, the word/topic prediction specifying a probability of the word being indicative of a second topic; and determining a document/topic prediction for the document based upon the feature/topic prediction and the word/topic prediction, the document/topic prediction specifying a probability of the document being indicative of a third topic.