Recommendations based on branding

Patent metadata

Field	Value
Publication number	US-9443209-B2
Application number	US-70761810-A
Country	US
Kind code	B2
Filing date	Feb 17, 2010
Priority date	Apr 30, 2009
Publication date	Sep 13, 2016
Grant date	Sep 13, 2016

Abstract

A method and a system for providing recommendations based on branding are disclosed. For example, a brand preference corresponding to a first brand and a first category may be identified based on user activity. A recommendation is provided to the user based on the brand preference. The recommendation may be provided based on a predetermined brand relationship comprising the first brand associated with the first category, a second brand associated with a second category, and a recommendation score between the first and second categories and brands. The recommendation may provided by accessing a relationships database to determine at least one brand relationship of the brand relationships corresponding to the brand preference.

First claim

What is claimed is:

1. A system comprising:
a memory to store an index comprising a plurality of brand relationships based on an analysis of a collection of user queries, each of the brand relationships comprising a first brand-category pair of a first brand and a first category, a second brand-category pair of a second brand and a second category, and a recommendation score calculated between the first brand-category pair and the second brand-category pair, the recommendation score of each brand relationship determined from user sessions having queries of both the first brand and first category and the second brand and second category;
a processor to implement a recommendation module to provide a recommendation based on an identified brand preference from an activity of a user by querying the index in the memory using the first brand and first category of the identified brand preference to identify the second brand-category pair having a highest recommendation score in the index; and
an expansion module to expand a seed set of brands corresponding to a category by analyzing the collection of user queries to determine a new brand to add to the seed set, the analyzing including mining a corpus containing a plurality of user queries by evaluating user queries containing a disjunction of brand terms.

2. The system of claim 1 , further comprising a mapping module to identify the brand relationships based on user activity.

3. The system of claim 1 , wherein the expansion module is further to identify the brand relationship between the first brand and the second brand based on an analysis of the collection of user queries.

4. The system of claim 1 , wherein the analyzing further comprises:
creating a candidate list comprising terms that occur frequently with seed brands;
removing terms from the candidate list that occur less than a minimum number of co-occurrences; and
removing terms from the candidate list that occur less than a minimum occurrence percentage.

5. The system of claim 1 , wherein the expansion module is to identify at least one variation of the first brand.

6. The system of claim 1 , wherein the expansion module is to determine that the first brand and the second brand co-occur in a session associated with the collection of user queries.

7. The system of claim 1 , wherein the expansion module is to generate the recommendation score.

8. The system of claim 1 , further comprising a semantics module to determine a semantic similarity between user queries.

9. The system of claim 8 , wherein the semantic similarity is based on common terms in a plurality of user queries.

10. The system of claim 8 , wherein the semantic similarity is based on queries submitted by a user during a user session.

11. The system of claim 8 , wherein the semantic similarity is based on a plurality of completed transactions.

12. The system of claim 1 , further comprising a categories module to determine a category associated with an item for sale or a user query in an online publication system.

13. The system of claim 1 , wherein the first category and the second category of the recommendation correspond to a same category.

14. The system of claim 1 , wherein the first brand and the second brand of the recommendation correspond to a same brand.

15. The system of claim 1 , wherein the recommendation module is to identify the second brand-category pair based on a user activity used to identify the brand preference.

16. The system of claim 1 , wherein the recommendation module is to select a brand relationship from the plurality of brand relationships based on respective recommendation scores of the plurality of brand relationships.

17. A method comprising:
receiving a user activity performed by a user in an online publication system; identifying a brand preference comprising a first brand and a first category from the user activity;
accessing an index comprising predetermined brand relationships, each predetermined brand relationship including a first brand-category pair of the first brand and the first category, a second brand-category pair of a second brand and a second category, and a recommendation score calculated between the first brand-category pair and the second brand-category pair, the recommendation score of each brand relationship determined from user sessions having queries of both the first brand and first category and the second brand and second category;
determining, using a hardware processor, a recommendation to provide the user based on the brand preference by querying the index using the first brand and the first category of the brand preference to determine the second brand-category pair having a highest recommendation score in the index;
providing the recommendation to the user; and expanding a seed set of brands corresponding to a category by analyzing a collection of user queries to determine a new brand to add to the seed set, the analyzing including mining a corpus containing a plurality of user queries by evaluating user queries containing a disjunction of brand terms.

18. The method of claim 17 , further comprising:
mining the user queries to identify relationships between the user queries;
mapping the user queries to categories; and
generating the plurality of brand relationships by mapping the relationships between the user queries.

19. The method of claim 17 , wherein the user activity comprises a selection from the group consisting of a user query, a bid on an item for sale, a completed purchase of an item, and a view of an item for sale.

20. A non-transitory computer-readable storage medium having instructions embodied thereon, the instructions executable by a computer to cause the computer to perform operations comprising:
receiving a user activity performed by a user in an online publication system; identifying a brand preference comprising a first brand and a first category from the user activity;
accessing an index comprising predetermined brand relationships, each predetermined brand relationship including a first brand-category pair of the first brand and the first category, a second brand-category pair of a second brand and a second category, and a recommendation score calculated between the first brand-category pair and the second brand-category pair, the recommendation score of each brand relationship determined from user sessions having queries of both the first brand and first category and the second brand and second category;
determining, using a hardware processor, a recommendation to provide the user based on the brand preference by querying the index using the first brand and the first category of the brand preference to determine the second brand-category pair having a highest recommendation score in the index;
providing the recommendation to the user; and expanding a seed set of brands corresponding to a category by analyzing a collection of user queries to determine a new brand to add to the seed set, the analyzing including mining a corpus containing a plurality of user queries by evaluating user queries containing a disjunction of brand terms.

Assignees

• Parikh Nishith
• Sundaresan Neelakantan
• Paypal Inc

Inventors

• Parikh Nishith
• Sundaresan Neelakantan

Classifications

• G06Q30/0631 Primary

  Recommending goods or services · CPC title

• G06Q30/02

  Marketing; Price estimation or determination; Fundraising · CPC title

• G06Q10/04 Primary

  Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem" (market predictions or forecasting for commercial activities G06Q30/0202) · CPC title

• G06F16/35

  Clustering; Classification · CPC title

• G06F16/36

  Creation of semantic tools, e.g. ontology or thesauri · CPC title