Targeted marketing using dynamically varied recommendations
Techniques are disclosed for presenting recommended marketing content that can be dynamically varied (e.g., in real time or near-real time) based on the particular content of a web page that a user is currently reading, the content that the user has previously read, or a combination of these factors. For example, the current interactions of a user with content of a web page, such as scroll position or eye fixation, can be used to identify which portion or portions of the web page content the user is currently reading and which portions of the web page content the user is skimming or skipping over. Keywords can then be generated based on the content in those portions of the web page that the user is currently reading. Next, recommended marketing content corresponding to the keywords can be retrieved and presented to the user.
1. A computer-implemented method comprising:
receiving digital content via a web browser;
displaying, on a display device, a first portion of the digital content in the web browser;
detecting a plurality of eye fixation locations on the displayed digital content using an eye tracking device, each eye fixation location represented by (x, y) coordinates;
receiving, by a processor, from the eye tracking device, data representing the plurality of eye fixation locations;
identifying multiple pairs of two consecutive eye fixation locations having a difference in coordinates (Δx, Δy) such that Δx is greater than a horizontal threshold value and Δy is less than a vertical threshold value, wherein the horizontal and vertical threshold values are indicative of a user reading a line of text in the digital content;
identifying a second portion of the digital content that corresponds to the identified multiple pairs of eye fixation locations;
generating keywords associated with the identified second portion of the digital content;
retrieving recommended content corresponding to the keywords; and
presenting the recommended content via the web browser.
2. The method of claim 1 , further comprising:
generating additional keywords based on the digital content in its entirety;
retrieving additional recommended marketing content corresponding to the additional keywords; and
presenting a mix of the recommended content and the additional recommended content in relative proportion to a weight.
3. The method of claim 1 , further comprising:
identifying a prior interaction of the user with the digital content; and
identifying a third portion of the digital content that was previously displayed to the user based on the prior interaction,
wherein the third portion of the digital content was displayed to the user for a dwell time that is greater than a threshold dwell time, and
wherein the generated keywords are associated with the third portion of the digital content.
4. The method of claim 1 , further comprising receiving a current scroll position within the web browser, wherein the identified second portion of the digital content includes the current scroll position.
5. The method of claim 1 , wherein the second portion of the digital content has a concentration of eye fixation locations greater than a concentration threshold.
6. The method of claim 1 , wherein generating the keywords includes applying natural language processing to the identified second portion of the digital content.
7. The method of claim 1 , wherein the digital content is at least one of a web page, an electronic mail message, and an electronic document.
8. A system comprising:
a display;
an eye tracking device configured to determine where on the display a user's eyes are fixated, and to generate (x, y) coordinates based on such determination;
a storage; and
a processor operatively coupled to the storage, the processor configured to receive coordinates from the eye tracking device and execute instructions stored in the storage that when executed cause the processor to carry out a process comprising:
receiving digital content via a web browser;
displaying a first portion of the digital content in the web browser;
receiving data generated by the eye tracking device that represents a plurality of eye fixation locations on the displayed digital content, wherein the eye fixation locations are (a) associated with a fixation time within a specified range, and (b) within a second portion of the digital content;
generating keywords associated with the second portion of the digital content;
retrieving recommended marketing content corresponding to the keywords; and
presenting the recommended marketing content via the web browser.
9. The system of claim 8 , wherein the process includes:
generating additional keywords based on the digital content in its entirety;
retrieving additional recommended marketing content corresponding to the additional keywords; and
presenting a mix of the recommended marketing content and the additional recommended marketing content in relative proportion to a weight.
10. The system of claim 8 , wherein the process includes:
identifying a prior interaction of a user with the digital content; and
identifying a third portion of the digital content that was previously displayed to the user based on the prior interaction,
wherein the third portion of the digital content was displayed to the user for a dwell time that is greater than a threshold dwell time, and
wherein the generated keywords are associated with the third portion of the digital content.
11. The system of claim 8 , wherein the process includes receiving a current scroll position within the web browser, and wherein the second portion of the digital content includes the current scroll position.
12. The system of claim 8 , wherein the specified range is between 100 milliseconds and 500 milliseconds.
13. The system of claim 8 , wherein generating the keywords includes applying natural language processing to the second portion of the digital content.
14. A non-transitory computer program product having instructions encoded thereon that when executed by one or more processors cause a process to be carried out, the process comprising:
receiving digital content via a web browser;
displaying, on a display device, a first portion of the digital content in the web browser;
detecting a plurality of eye tracking positions within the web browser using an eye tracking device, each eye tracking position represented by (x, coordinates;
receiving, from the eye tracking device, (x, y) coordinates of a first eye tracking position that defines the first portion of the digital content, wherein the first portion of the digital content is displayed for a first dwell time;
receiving data representing a current interaction of a user with the digital content via the web browser, wherein the current interaction is a scrolling operation;
receiving, from the eye tracking device, (x, y) coordinates of a second eye tracking position that defines a second portion of the digital content, wherein the second portion of the digital content is displayed for a second dwell time;
identifying a portion of the digital content that is displayed for a dwell time that is greater than a threshold dwell time;
generating keywords associated with the identified portion of the digital content;
retrieving recommended content corresponding to the keywords; and
presenting the recommended content via the web browser.
15. The non-transitory computer program product of claim 14 , wherein the process includes:
generating additional keywords based on the digital content in its entirety;
retrieving additional recommended content corresponding to the additional keywords; and
presenting a mix of the recommended content and the additional recommended content in relative proportion to a weight.
16. The non-transitory computer program product of claim 14 , wherein the process includes:
identifying a prior interaction of the user with the digital content; and
identifying a third portion of the digital content that was previously displayed to the user based on the prior interaction,
wherein the third portion of the digital content was displayed to the user for a third dwell time that is greater than the threshold dwell time, and
wherein the generated keywords are associated with the third portion of the digital content.
17. The non-transitory computer program product of claim 14 , wherein the process includes receiving a current scroll position within the web browser, and wherein at least one of the first and second portions of the digital content includes the current scroll position.
18. The non-transitory computer program product of claim 14 , wherein the first and second eye tracking positions are each associated with a fixation time between 100 milliseconds and 500 milliseconds.
19. The non-transitory computer program product of claim 14 , wherein generating the keywords includes applying natural language processing to the identified portion of the digital content.
20. The non-transitory computer program product of claim 14 , wherein the digital content is at least one of a web page, an electronic mail message, and an electronic document.