Influence maximization with viral product design
View Patent ↗The disclosure includes use of a feature-aware propagation model to identify one or more features of a product and one or more person(s), or members of a social network, to target, or user, for marketing the product having the identified features. The one or more person(s) identified using the model may be the person(s), or member(s), of a social network determined to have a maximum capability, relative to other members of the social network, for influencing the members of the social network in adopting, e.g., purchasing, a product having the identified features. In addition, parameters of the model may be determined using information about the social network, user preferences, and the products and features of the products.
1. A method comprising:
using, by at least one computing system, a feature-aware propagation model having a set of model parameters in selecting a seed set of users that are members of a social network for influencing other users of the social network in combination with selecting, by the feature-aware propagation model, a product feature set comprising a number of features of a product, the feature-aware propagation model further determining a likelihood of the combination to maximize adoption of the product having the product's product feature set by members of the social network; and
identifying, by the at least one computing system, the combination of the user seed set of users and the product feature set for the product selected using the feature-aware propagation model, the user seed set comprising a number of users for influencing other users with respect to the product with its product feature set, the identified combination of the seed set of users that are members of a social network for influencing other users of the social network and the product's product feature set having the determined likelihood to maximize adoption of the product having the product's product feature set by members of the social network that is greater than the determined likelihood corresponding to at least one other combination of user seed set and product feature set considered by the feature-aware propagation model.
2. The method of claim 1 , the set of model parameters comprising an influence strength for each of a plurality of user pairings, a part-worth utility for each of a plurality of pairings of a user of the social network and a feature of the product, and a hurdle for each of a plurality of users of the social network, the influence strength proving a measure of a strength of influence of a first user in a user pairing over a second user in the user pairing, the part-worth utility providing a measure of importance of a given product feature to a given user, and a hurdle providing a product adoption equilibrium measure for the given user.
3. The method of claim 2 , further comprising:
generating, by the at least one computing system, the set of model parameters using input comprising a social graph representing the social network, information identifying a set of products and a set of features for each product of the set of products and feature preferences and information identifying product and feature preferences of members of the social network.
4. The method of claim 1 , the identifying the combination of the user seed and product feature sets determined to have a greater likelihood to maximize adoption of the product having the product feature set by members of the social network further comprising:
performing a plurality of iterations, in each iteration updating one of the user seed set and the product feature set;
in an iteration that updates the user seed set:
determining a spread of influence measure for each user not already a member of the user seed set; and
updating the user seed set to include the user not already a member of the user seed set that has the greatest determined spread of influence measure relative to the spread of influence measures determined for the other users not already a member of the user seed set;
in an iteration that updates the product feature set:
determining the spread of influence measure for each of a plurality of product feature updates; and
updating the product feature set using the product feature update of the plurality that has the greatest determined spread of influence measure relative to the spread of influence measures determined for the other product feature updates of the plurality.
5. The method of claim 4 , further comprising:
determining the spread of influence measure for each of a plurality of product feature updates comprising, for each product feature not in the product feature set, temporarily adding the product feature to the product feature set and determining a spread of influence measure for the product feature set including the product feature; and
updating the product feature set by adding the product feature having the greatest determined spread of influence measure relative to the spread of influence measures determined for one or more other product features temporarily added to the product feature set.
6. The method of claim 4 , further comprising:
determining the spread of influence measure for each of a plurality of product feature updates comprising, for each product feature in the product feature set, temporarily removing the product feature from the product feature set and determining a spread of influence measure for the product feature set excluding the removed product feature; and
updating the product feature set by removing the product feature having the greatest determined spread of influence measure relative to the spread of influence measures determined for one or more other product features temporarily removed from the product feature set.
7. The method of claim 4 , further comprising:
in the iteration that updates the product feature set:
determining a spread of influence measure for each of a plurality of alternate user seed sets, each alternate user seed set comprising a number of users equal to the number of users in the user seed set and comprising a set of users selected from a subset of the social network excluding those one or more users in the user seed set;
comparing the spread of influence measure for the user seed set with the spread of influence measure determined for each of the plurality of alternate user seed sets; and
replacing the user seed set with one of the plurality of alternate user seed sets if it is determined that the one of the plurality of alternate user seed sets has an associated spread of influence measure that is greater than the user seed set's spread of influence measure.
8. A system comprising:
at least one computing device, each computing device comprising a processor and a storage medium for tangibly storing thereon program logic for execution by the processor, the stored program logic comprising:
using logic executed by the processor for using a feature-aware propagation model having a set of model parameters in selecting a seed set of users that are members of a social network for influencing other users of the social network in combination with selecting, by the feature-aware propagation model, a product feature set comprising a number of features of a product, the feature-aware propagation model further determining a likelihood of the combination to maximize adoption of the product having the product's product feature set by members of the social network; and
identifying logic executed by the processor for identifying the combination of the user seed set of users and the product feature set for the product selected using the feature-aware propagation model, the user seed set comprising a number of users for influencing other users with respect to the product with its product feature set, the identified combination of the seed set of users that are members of a social network for influencing other users of the social network and the product's product feature set having the determined likelihood to maximize adoption of the product having the product's product feature set by members of the social network that is greater than the determined likelihood corresponding to at least one other combination of user seed set and product feature set considered by the feature-aware propagation model.
9. The system of claim 8 , the set of model parameters comprising an influence strength for each of a plurality of user pairings, a part-worth utility for each of a plurality of pairings of a user of the social network and a feature of the product, and a hurdle for each of a plurality of users of the social network, the influence strength proving a measure of a strength of influence of a first user in a user pairing over a second user in the user pairing, the part-worth utility providing a measure of importance of a given product feature to a given user, and a hurdle providing a product adoption equilibrium measure for the given user.
10. The system of claim 9 , the stored program logic further comprising:
generating logic executed by the processor for generating the set of model parameters using input comprising a social graph representing the social network, information identifying a set of products and a set of features for each product of the set of products and information identifying product and feature preferences of members of the social network.
11. The system of claim 8 , the identifying logic executed by the processor for identifying the combination of the user seed and product feature sets determined to have a greater likelihood to maximize adoption of the product having the product feature set by members of the social network further comprising:
performing logic executed by the processor for performing a plurality of iterations, in each iteration updating one of the user seed set and the product feature set;
in an iteration that updates the user seed set, the stored program logic further comprising:
determining logic executed by the processor for determining a spread of influence measure for each user not already a member of the user seed set; and
updating logic executed by the processor for updating the user seed set to include the user not already a member of the user seed set that has the greatest determined spread of influence measure relative to the spread of influence measures determined for the other users not already a member of the user seed set;
in an iteration that updates the product feature set, the stored program logic-further comprising:
determining logic executed by the processor for determining the spread of influence measure for each of a plurality of product feature updates; and
updating logic executed by the processor for updating the product feature set using the product feature update of the plurality that has the greatest determined spread of influence measure relative to the spread of influence measures determined for the other product feature updates of the plurality.
12. The system of claim 11 , the stored program logic further comprising:
determining logic executed by the processor for determining the spread of influence measure for each of a plurality of product feature updates by, for each product feature not in the product feature set, temporarily adding the product feature to the product feature set and determine a spread of influence measure for the product feature set including the product feature; and
updating logic executed by the processor for updating the product feature set by adding the product feature having the greatest determined spread of influence measure relative to the spread of influence measures determined for one or more other product features temporarily added to the product feature set.
13. The system of claim 11 , the stored program logic further comprising:
determining logic executed by the processor for determining the spread of influence measure for each of a plurality of product feature updates comprising, for each product feature in the product feature set, temporarily removing the product feature from the product feature set and determining a spread of influence measure for the product feature set excluding the removed product feature; and
updating logic executed by the processor for updating the product feature set by removing the product feature having the greatest determined spread of influence measure relative to the spread of influence measures determined for one or more other product features temporarily removed from the product feature set.
14. The system of claim 11 , the stored program logic further comprising:
in the iteration that updates the product feature set:
determining logic executed by the processor for determining a spread of influence measure for each of a plurality of alternate user seed sets, each alternate user seed set comprising a number of users equal to the number of users in the user seed set and comprising a set of users selected from a subset of the social network excluding those one or more users in the user seed set;
comparing logic executed by the processor for comparing the spread of influence measure for the user seed set with the spread of influence measure determined for each of the plurality of alternate user seed sets; and
replacing logic executed by the processor for replacing the user seed set with one of the plurality of alternate user seed sets if it is determined that the one of the plurality of alternate user seed sets has an associated spread of influence measure that is greater than the user seed set's spread of influence measure.
15. A computer readable non-transitory storage medium for tangibly storing thereon computer readable instructions that when executed cause at least one processor to:
use a feature-aware propagation model having a set of model parameters in selecting a seed set of users that are members of a social network for influencing other users of the social network in combination with selecting, by the feature-aware propagation model, a product feature set comprising a number of features of a product, the feature-aware propagation model further determining a likelihood of the combination to maximize adoption of the product having the product's product feature set by members of the social network; and
identify the combination of the user seed set of users and the product feature set for the product selected using the feature-aware propagation model, the user seed set comprising a number of users for influencing other users with respect to the product with its product feature set, the identified combination of the seed set of users that are members of a social network for influencing other users of the social network and the product's product feature set having the determined likelihood to maximize adoption of the product having the product's product feature set by members of the social network that is greater than the determined likelihood corresponding to at least one other combination of user seed set and product feature set considered by the feature-aware propagation model.
16. The computer readable non-transitory storage medium of claim 15 , the set of model parameters comprising an influence strength for each of a plurality of user pairings, a part-worth utility for each of a plurality of pairings of a user of the social network and a feature of the product, and a hurdle for each of a plurality of users of the social network, the influence strength proving a measure of a strength of influence of a first user in a user pairing over a second user in the user pairing, the part-worth utility providing a measure of importance of a given product feature to a given user, and a hurdle providing a product adoption equilibrium measure for the given user.
17. The computer readable non-transitory storage medium of claim 16 , the instructions further comprising instructions to:
generate the set of model parameters using input comprising a social graph representing the social network, information identifying a set of products and a set of features for each product of the set of products and information identifying product and feature preferences of members of the social network.
18. The computer readable non-transitory storage medium of claim 15 , the instructions to the combination of the user seed and product feature sets determined to have a greater likelihood to maximize adoption of the product having the product feature set by members of the social network further comprising instructions to:
perform a plurality of iterations, in each iteration updating one of the user seed set and the product feature set;
in an iteration that updates the user seed set, the instructions further comprising instructions to:
determine a spread of influence measure for each user not already a member of the user seed set; and
update the user seed set to include the user not already a member of the user seed set that has the greatest determined spread of influence measure relative to the spread of influence measures determined for the other users not already a member of the user seed set;
in an iteration that updates the product feature set, the instructions further comprising instructions to:
determine the spread of influence measure for each of a plurality of product feature updates; and
update the product feature set using the product feature update of the plurality that has the greatest determined spread of influence measure relative to the spread of influence measures determined for the other product feature updates of the plurality.
19. The computer readable non-transitory storage medium of claim 18 , the instructions further comprising instructions to:
determine the spread of influence measure for each of a plurality of product feature updates by, for each product feature not in the product feature set, temporarily adding the product feature to the product feature set and determine a spread of influence measure for the product feature set including the product feature; and
update the product feature set by adding the product feature having the greatest determined spread of influence measure relative to the spread of influence measures determined for one or more other product features temporarily added to the product feature set.
20. The computer readable non-transitory storage medium of claim 18 , the instructions further comprising instructions to:
determine the spread of influence measure for each of a plurality of product feature updates comprising, for each product feature in the product feature set, temporarily removing the product feature from the product feature set and determining a spread of influence measure for the product feature set excluding the removed product feature; and
update the product feature set by removing the product feature having the greatest determined spread of influence measure relative to the spread of influence measures determined for one or more other product features temporarily removed from the product feature set.
21. The computer readable non-transitory storage medium of claim 18 , the instructions further comprising instructions to:
in the iteration that updates the product feature set:
determine a spread of influence measure for each of a plurality of alternate user seed sets, each alternate user seed set comprising a number of users equal to the number of users in the user seed set and comprising a set of users selected from a subset of the social network excluding those one or more users in the user seed set;
compare the spread of influence measure for the user seed set with the spread of influence measure determined for each of the plurality of alternate user seed sets; and
replace the user seed set with one of the plurality of alternate user seed sets if it is determined that the one of the plurality of alternate user seed sets has an associated spread of influence measure that is greater than the user seed set's spread of influence measure.