Implicit question query identification

• US 9,898,554 B2
• Filed: 11/18/2013
• Issued: 02/20/2018
• Est. Priority Date: 11/18/2013
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Gupta, Nitin, Popat, Preyas, Baker, Steven D., Venkatachary, Srinivasan
• Primary Examiner(s): Ho, Binh V
• Application Number:
  US14/083,057
• Publication Number:
  US 20150142851A1
• Time in Patent Office: 1,555 Days
• Field of Search: 707779
• CPC Class Codes: G06F 16/9032 Query formulation

Abstract

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for identifying implicit question queries. In one aspect, a method includes receiving a query in unstructured form, comparing terms of the query to query templates, determining, based on the comparison, a match of the query terms to a first query template, wherein the first query template is not determined to be indicative of a question query, determining, based on the first query template, a second query template, and determining that the query is an implicit question query in response to the second query template being indicative of a question queries.

Claims

1. A method implemented in a data processing apparatus, the method comprising:

receiving, at the data processing apparatus, a query in unstructured form, the query including a plurality of terms, each of the terms at a respective ordinal position in the query;

comparing, by the data processing apparatus, the terms of the query to a plurality of query templates, each query template including terms of variable data and terms of invariable data, wherein the terms of variable data specifies a class of terms and the terms of invariable data specifies particular instances of terms;

determining, by the data processing apparatus and based on the comparison, a match of the query terms to a first query template, wherein the first query template is not determined to be indicative of a question query;

determining, by the data processing apparatus and based on the first query template, a second query template; and

determining, by the data processing apparatus, that the query is an implicit question query in response to the second query template being indicative of a question query;

wherein determining, based on the first query template, the second query template comprises;

generating synthetic templates from the first template, each synthetic template being the first template in combination with one of a corresponding plurality of question prefix terms; and

determining that one of the synthetic templates specify the second query template in the plurality of query templates.

2. The method of claim 1, further comprising validating the second query template based on query statistics of queries that match the second query template.

3. The method of claim 2, wherein validating the second query template comprises validating the second query template based on one or more of a unique query count statistic, a unique entity count statistic, a query click count statistic, a unique entity recall statistic, and a template variation statistic.

4. A method implemented in a data processing apparatus, the method comprising:

receiving, at the data processing apparatus, a query in unstructured form, the query including a plurality of terms, each of the terms at a respective ordinal position in the query;

comparing, by the data processing apparatus, the terms of the query to a plurality of query templates, each query template including terms of variable data and terms of invariable data, wherein the terms of variable data specifies a class of terms and the terms of invariable data specifies particular instances of terms;

determining, by the data processing apparatus and based on the comparison, a match of the query terms to a first query template, wherein the first query template is not determined to be indicative of a question query;

determining, by the data processing apparatus and based on the first query template, a second query template; and

determining, by the data processing apparatus, that the query is an implicit question query in response to the second query template being indicative of a question query;

wherein;

comparing the terms of the query comprises selecting proper subsets of terms of the query, each proper subset being sequentially ordinal terms and being no longer than k terms; and

comparing the terms of the query to a plurality of query templates comprises comparing each proper subset of query terms to the plurality of query templates.

5. A method implemented in a data processing apparatus, the method comprising:

receiving, at the data processing apparatus, a query in unstructured form, the query including a plurality of terms, each of the terms at a respective ordinal position in the query;

comparing, by the data processing apparatus, the terms of the query to a plurality of query templates, each query template including terms of variable data and terms of invariable data, wherein the terms of variable data specifies a class of terms and the terms of invariable data specifies particular instances of terms;

determining, by the data processing apparatus and based on the comparison, a match of the query terms to a first query template, wherein the first query template is not determined to be indicative of a question query;

determining, by the data processing apparatus and based on the first query template, a second query template; and

determining, by the data processing apparatus, that the query is an implicit question query in response to the second query template being indicative of a question query;

wherein;

comparing the terms of the query comprises;

determining a root subject term of the query;

determining sibling phrases of two or more terms to the root subject term, each sibling phrase not including the root subject term and not including terms common to other sibling phrases;

pruning one or more sibling phrases to form a pruned query of the root subject term and at least one sibling phrase; and

comparing the terms of the query to a plurality of query templates comprises comparing the terms of the pruned query to the plurality of query templates.

6. A system, comprising:

one or more data processing apparatus; and

one or more data stores in communication with the one or more data processing apparatus and storing instructions that upon execution cause the one or more data processing apparatus to perform operations comprising;

receiving a query in unstructured form, the query including a plurality of terms, each of the terms at a respective ordinal position in the query;

comparing the terms of the query to a plurality of query templates, each query template including terms of variable data and terms of invariable data, wherein the terms of variable data specifies a class of terms and the terms of invariable data specifies particular instances of terms;

determining, based on the comparison, a match of the query terms to a first query template, wherein the first query template is not determined to be indicative of a question query;

determining, based on the first query template, a second query template; and

determining that the query is an implicit question query in response to the second query template being indicative of a question query;

wherein determining, based on the first query template, the second query template comprises;

generating synthetic templates from the first template, each synthetic template being the first template in combination with one of a corresponding plurality of question prefix terms; and

determining that one of the synthetic templates specify the second query template in the plurality of query templates.

7. The system of claim 6, wherein the operations further comprise validating the second query template based on query statistics of queries that match the second query template.

8. The system of claim 7, wherein validating the second query template comprises validating the second query template based on one or more of a unique query count statistic, a unique entity count statistic, a query click count statistic, a unique entity recall statistic, and a template variation statistic.

9. A system, comprising:

one or more data processing apparatus; and

one or more data stores in communication with the one or more data processing apparatus and storing instructions that upon execution cause the one or more data processing apparatus to perform operations comprising;

receiving a query in unstructured form, the query including a plurality of terms, each of the terms at a respective ordinal position in the query;

comparing the terms of the query to a plurality of query templates, each query template including terms of variable data and terms of invariable data, wherein the terms of variable data specifies a class of terms and the terms of invariable data specifies particular instances of terms;

determining, based on the comparison, a match of the query terms to a first query template, wherein the first query template is not determined to be indicative of a question query;

determining, based on the first query template, a second query template; and

determining that the query is an implicit question query in response to the second query template being indicative of a question query;

wherein;

comparing the terms of the query comprises selecting proper subsets of terms of the query, each proper subset being sequentially ordinal terms and being no longer than k terms; and

comparing the terms of the query to a plurality of query templates comprises comparing each proper subset of query terms to the plurality of query templates.

10. A system, comprising:

one or more data processing apparatus; and

one or more data stores in communication with the one or more data processing apparatus and storing instructions that upon execution cause the one or more data processing apparatus to perform operations comprising;

receiving a query in unstructured form, the query including a plurality of terms, each of the terms at a respective ordinal position in the query;

comparing the terms of the query to a plurality of query templates, each query template including terms of variable data and terms of invariable data, wherein the terms of variable data specifies a class of terms and the terms of invariable data specifies particular instances of terms;

determining, based on the comparison, a match of the query terms to a first query template, wherein the first query template is not determined to be indicative of a question query;

determining, based on the first query template, a second query template; and

determining that the query is an implicit question query in response to the second query template being indicative of a question query;

wherein;

comparing the terms of the query comprises;

determining a root subject term of the query;

determining sibling phrases of two or more terms to the root subject term, each sibling phrase not including the root subject term and not including terms common to other sibling phrases;

pruning one or more sibling phrases to form a pruned query of the root subject term and at least one sibling phrase; and

comparing the terms of the query to a plurality of query templates comprises comparing the terms of the pruned query to the plurality of query templates.