Searching questions based on topic and focus
AssigneeMicrosoft Corporation
InventorsYunbo Cao, Chin-Yew Lin
Life of this patent
Estimating the term…
Abstract
A method and system for determining the relevance of questions to a queried question based on topics and focuses of the questions is provided. A question search system provides a collection of questions with topics and focuses. Upon receiving a queried question, the question search system identifies a queried topic and queried focus of the queried question. The question search system generates a score indicating the relevance of a question of the collection to the queried question based on a language model of the topic of the question and a language model of the focus of the question.
Claims20
A method in a computing device for identifying questions relevant to a queried question, the method comprising:
- providing a collection of questions having terms, each question having a topic of one or more terms of the question and a focus of one or more terms of the question;
- for each topic and for each term in the questions of the collection, calculating by the computing device a probability of generating that term from a language model of that topic;
- for each focus and for each term in the questions of the collection, calculating by the computing device a probability of generating that term from a language model of that focus;
- receiving a queried question having terms;
- identifying a queried topic and a queried focus of the queried question;
- for each of a plurality of questions of the collection,
- calculating a topic probability of the queried topic as a probability of generating the terms of the queried topic from a language model of the topic of the question;
- calculating a focus probability of the queried focus as a probability of generating the terms of the queried focus from a language model of the focus of the question;
- and
- generating the probability of the queried question from a language model of the question using the calculated topic probability and the calculated focus probability of the question, the probability indicating the relevance of the question to the queried question.
A computer-readable storage medium containing instructions for controlling a computing device to identify questions related to a queried question, by a method comprising:
- providing a collection of questions, each question having terms, each term having a topic and a focus, the topics and focuses of the questions of the collection are identified by generating a question tree of the questions of the collection and generating a cut of the question tree, a topic of a question comprising the terms of the question before the cut and the focus of a question comprising the terms of the question after the cut;
- for each topic and for each term in the questions of the collection, calculating a probability of generating that term from a language model of the topic;
- for each focus and for each term in the questions of the collection, calculating a probability of generating that term from a language model of the focus;
- receiving a queried question having terms;
- identifying a queried topic and a queried focus of the question;
- for each of a plurality of questions of the collection,
- calculating a topic probability of the queried topic that factors in a probability of generating the terms of the queried topic from a language model of the topic of the question, a collection probability of the terms occurring in the collection to account for sparseness of a term in the collection, and a translation probability that a term of the queried question is a translation of a term of a question;
- calculating a focus probability of the queried focus that factors in a probability of generating the terms of the queried focus from a language model of the focus of the question, a collection probability of the terms occurring in the collection to account for sparseness of a term in the collection, and a translation probability that a term of the queried question is a translation of a term of a question;
- and
- generating a probability that the queried question is from a language model of the question using a weighted combination of the calculated topic probability and the calculated focus probability of the question.
A computing device for determining the relevance of questions related to a queried question, comprising:
- a data store storing a collection of questions, each question having a topic of one or more terms and a focus of one or more terms;
- a memory storing computer-executable instructions of:
- a first component that receives a queried question having terms;
- a second component that identifies a queried topic and a queried focus of the queried question;
- and
- a third component that, for each of a plurality of questions of the collection, generates a score indicating the relevance of the question to the queried question using a language model of the topic of the question and a language model of the focus of the question;
- and
- a processor that executes the computer-executable instructions stored in the memory.
Dependencies are inferred from the claim text.