Modifying search result ranking based on implicit user feedback
The present disclosure includes systems and techniques relating to ranking search results of a search query. In general, the subject matter described in this specification can be embodied in a computer-implemented method that includes determining a measure of relevance for a document result within a context of a search query for which the document result is returned, the determining being based on a first number in relation to a second number, the first number corresponding to longer views of the document result, and the second number corresponding to at least shorter views of the document result; and outputting the measure of relevance to a ranking engine for ranking of search results, including the document result, for a new search corresponding to the search query. The subject matter described in this specification can also be embodied in various corresponding computer program products, apparatus and systems.
1. A system comprising:
one or more computers and one or more storage devices on which are stored instructions that are operable, when executed by the one or more computers, to cause the one or more computers to perform operations comprising:
determining a first number representing a count of longer views of a document following selections of search results, where the search results reference the document in response to submissions of a search query to a search engine, wherein the submissions of the search query comprise submissions by speech input into client devices, and wherein longer views of a document comprise audio presentations, on client devices, longer than a first threshold length of time;
determining a second number representing a count of shorter views of the document following the selections of the search results, where the search results reference the document in response to submissions of the search query, wherein shorter views of the document comprise audio presentations, on client devices, shorter than a second threshold length of time;
determining a measure of relevance for the document with respect to the search query based on a ratio between the first number and the second number; and
outputting the measure of relevance to a ranking engine for ranking of search results, including new search results referencing the document, for a new search responsive to a new submission of the search query.
2. The system of claim 1 , wherein determining the measure of relevance is further based on dividing the first number by a count of total views of the document.
3. The system of claim 1 , the operations further comprising:
tracking individual selections of the search results referencing the document within a context of the search query for which the search results are returned;
weighting document views resulting from the selections based on viewing length information to produce weighted views of the document, the viewing length information including listening length information for audio presentation; and
combining the weighted views of the document to determine the first number.
4. The system of claim 3 , wherein determining the measure of relevance comprises dividing the first number by the second number.
5. The system of claim 3 , wherein the weighting comprises applying a continuous function to the document views resulting from the selections.
6. The system of claim 3 , wherein the weighting comprises applying a discontinuous function to the document views resulting from the selections.
7. The system of claim 6 , wherein applying the discontinuous function comprises:
classifying the individual selections of the document into at least three viewing time categories, the times including presentation times for audio presentations; and
assigning weights to the individual selections based on results of the classifying.
8. The system of claim 3 , wherein the weighting comprises weighting the document views based on the viewing length information in conjunction with a viewing length differentiator.
9. The system of claim 8 , wherein the viewing length differentiator includes a factor governed by a determined category of the search query, and the weighting comprises weighting the document views based on the determined category of the search query.
10. The system of claim 8 , wherein the viewing length differentiator includes a factor governed by a determined type of a user generating the individual selections, and the weighting comprises weighting the document views based on the determined type of the user.
11. A computer-implemented method comprising:
determining a first number representing a count of longer views of a document following selections of search results, where the search results reference the document in response to submissions of a search query to a search engine, wherein the submissions of the search query comprise submissions by speech input into client devices, and wherein longer views of a document comprise audio presentations, on client devices, longer than a first threshold length of time;
determining a second number representing a count of shorter views of the document following the selections of the search results, where the search results reference the document in response to submissions of the search query, wherein shorter views of the document comprise audio presentations, on client devices, shorter than a second threshold length of time;
determining a measure of relevance for the document with respect to the search query based on a ratio between the first number and the second number; and
outputting the measure of relevance to a ranking engine for ranking of search results, including new search results referencing the document, for a new search responsive to a new submission of the search query.
12. The method of claim 11 , wherein determining the measure of relevance is further based on dividing the first number by a count of total views of the document.
13. The method of claim 11 , further comprising:
tracking individual selections of the search results referencing the document within a context of the search query for which the search results are returned;
weighting document views resulting from the selections based on viewing length information to produce weighted views of the document, the viewing length information including listening length information for audio presentation; and
combining the weighted views of the document to determine the first number.
14. The method of claim 13 , wherein determining the measure of relevance comprises dividing the first number by the second number.
15. The method of claim 13 , wherein the weighting comprises applying a continuous function to the document views resulting from the selections.
16. The method of claim 13 , wherein the weighting comprises applying a discontinuous function to the document views resulting from the selections.
17. The method of claim 16 , wherein applying the discontinuous function comprises:
classifying the individual selections of the document into at least three viewing time categories, the times including presentation times for audio presentations; and
assigning weights to the individual selections based on results of the classifying.
18. The method of claim 13 , wherein the weighting comprises weighting the document views based on the viewing length information in conjunction with a viewing length differentiator.
19. The method of claim 18 , wherein the viewing length differentiator includes a factor governed by a determined category of the search query, and the weighting comprises weighting the document views based on the determined category of the search query.
20. The method of claim 18 , wherein the viewing length differentiator includes a factor governed by a determined type of a user generating the individual selections, and the weighting comprises weighting the document views based on the determined type of the user.
21. A computer program product, encoded on a computer-readable medium, operable to cause data processing apparatus to perform operations comprising:
determining a first number representing a count of longer views of a document following selections of search results, where the search results reference the document in response to submissions of a search query to a search engine, wherein the submissions of the search query comprise submissions by speech input into client devices, and wherein longer views of a document comprise audio presentations, on client devices, longer than a first threshold length of time;
determining a second number representing a count of shorter views of the document following the selections of the search results, where the search results reference the document in response to submissions of the search query, wherein shorter views of the document comprise audio presentations, on client devices, shorter than a second threshold length of time;
determining a measure of relevance for the document with respect to the search query based on a ratio between the first number and the second number; and
outputting the measure of relevance to a ranking engine for ranking of search results, including new search results referencing the document, for a new search responsive to a new submission of the search query.