Real-time detection of intent-less engagements in digital content distribution systems

1. A computer-implemented method comprising:

for each given campaign of a first plurality of campaigns, monitoring engagements and conversions of the given campaign when presented in a plurality of traffic segments, wherein each of the plurality of traffic segments comprises traffic related to a plurality of users;

calculating an aggregated conversion rate of each given campaign of the plurality of campaigns when presented in the plurality of traffic segments, said calculating the aggregated conversion rate of each given campaign comprising:

obtaining a first sum of observed conversions of the given campaign when presented in the plurality of traffic segments;

obtaining a second sum of observed engagements of the given campaign when presented in the plurality of traffic segments; and

calculating the aggregated conversion rate of the given campaign based on the first sum and the second sum, wherein the aggregated conversion rate indicates a probability of a conversion to occur for an engagement with the given campaign when presented in any of the plurality of traffic segments; and

for a given traffic segment presenting one or more campaigns of the plurality of campaigns, calculating an estimated quality score of the given traffic segment, said calculating the estimated quality score of the given traffic segment comprising:

determining a third sum of observed conversions in the one or more campaigns when presented in the given traffic segment;

determining an expected sum of conversions of the one or more campaigns based on aggregated conversion rates of the one or more campaigns, as calculated in said calculating the aggregated conversion rate of each given campaign, and based on a number of observed engagements with each one of the one or more campaigns when presented in the given traffic segment; and

based on the third sum of observed conversions and the expected sum of conversions in the given traffic segment, calculating the estimated quality score of the given traffic segment, wherein the estimated quality score of the given traffic segment is EQS being indicative of a quality level of the observed engagements with the each one of the one or more campaigns when presented in the given traffic segment, wherein a quality level that is below a threshold indicates that the observed engagements comprise at least some intent-less engagements or fraudulent engagements in the given traffic segment.

2. The method of claim 1 , wherein each traffic segment from the plurality of traffic segments is associated with a different content publisher.

3. The method of claim 1 , wherein said calculating the estimated quality score further comprises:

assigning a respective weight to each campaign of the one or more campaigns;

summing in the third sum a weighted value of the observed conversions determined in the one or more campaigns in the given traffic segment, according to the respective weight;

summing a weighted value of the expected sum of conversions of the one or more campaigns according to the respective weight, to thereby obtain a segment-specific expected conversions value; and

calculating the estimated quality score for the given traffic segment, based on a ratio between the third sum and the segment-specific expected conversions value.

4. The method of claim 1 further comprising, performing at least one of (i) and (ii):

i. determining that an association exists between first and second campaigns of the plurality of campaigns; and

calculating an aggregated conversion rate of the first campaign using data that includes observed conversions and observed engagements of Vie second campaign,

ii. determining, that an association exists between first and second traffic segments of the plurality of traffic segments; and

calculating a segment-specific estimated quality score of the first traffic segment using data that includes observed conversions and expected conversions Of the second traffic segment.

5. The method of claim 1 further comprising:

for a specific campaign from the one or more campaigns, calculating an adaptive estimated quality score of the specific campaign indicating an estimated quality score of the specific campaign when presented in the given traffic segment, said calculating the adaptive estimated quality score comprising:

determining a fourth sum of observed engagements with the specific campaign when presented in the given traffic segment;

determining a fifth sum of observed conversions in the specific campaign when presented in the given traffic segment;

calculating a pair-specific quality score of the specific campaign based on a ratio between the fourth sum and the fifth sum;

calculating the adaptive estimated quality score of the specific campaign based on a combination of a weighted value of the estimated quality score of the given traffic segment and a weighted value of the pair-specific quality score.

6. The method of claim 1 further comprising at least one of:

i. assigning a respective first weight to each engagement;

ii. assigning a respective second weight to each conversion; and

iii. assigning a respective third weight to each campaign.

7. The method of claim 1 further comprising:

in the event that the quality level is below the threshold, generating an alert.

8. The method claim 1 further comprising: determining the estimated quality score of the given traffic segment based on a ratio between the third sum of observed conversions and the expected sum of conversions in the given traffic segment.

9. The method of claim 1 further comprising: upon determining that the observed engagements comprise the at least some intent-less engagements or fraudulent engagements, performing one or more of:

generating an alert indicating a suspicion of an engagement fraud; and

executing a preventive action directed for mitigating the engagement fraud.

10. The method of claim 1 , wherein a reward is given per a user engagement with content of the one or more campaigns presented in the given traffic segment, the method further comprising:

automatically updating a value of the reward according to the estimated quality score of the given traffic segment, thereby mitigating intent-less engagements.

11. The method of any one of the preceding claims, wherein the given traffic segment is associated to any one of: a website; a time frame; and a group of users having one or more common characteristics.

12. A computerized device comprising one or more computer processors configured and operable to:

for each given campaign of a plurality of campaigns, obtain information pertaining to engagements and conversions of the given campaign when presented in a plurality of traffic segments, wherein each of the plurality of traffic segments comprises traffic related to a plurality of users; and

calculate an aggregated conversion rate of each given campaign of the plurality of campaigns when presented in the plurality of traffic segments, said calculate the aggregated conversion rate of each given campaign comprising:

obtain a first sum of observed conversions of the given campaign when presented in the plurality of traffic segments;

obtain a second sum of observed engagements of the given campaign when presented in the plurality of traffic segments; and

calculate the aggregated conversion rate of the given campaign based on the first sum and the second sum, wherein the aggregated conversion rate indicates a probability of a conversion to occur for an engagement with the given campaign when presented in any of the plurality of traffic segments; and

for a given traffic segment presenting one or more campaigns of the plurality of campaigns, calculate an estimated quality score of the given traffic segment, said calculate the estimated quality score of the given traffic segment comprising:

determine a third sum of observed conversions in the one or more campaigns when presented in the given traffic segment;

determine an expected sum of conversions of the one or more campaigns based on aggregated conversion rates of the one or more campaigns, as calculated in said calculating the aggregated conversion rate of each given campaign, and based on a number of observed engagements with each one of the one or more campaigns when presented in the given traffic segment; and

based on the third sum of observed conversions and the expected sum of conversions in the given traffic segment, calculate the estimated quality score of the given traffic segment, wherein the estimated quality score of the given traffic segment is indicative of a quality level of the observed engagements with the each one of the one or more campaigns when presented in the given traffic segment, wherein a quality level that is below a threshold indicates that the observed engagements comprise at least some intent-less engagements or fraudulent engagements in the given traffic segment.

13. The computerized device of claim 12 , wherein said calculate the estimated quality score further comprises:

assigning a respective weight to each campaign of the one or more campaigns;

summing in the third sum a weighted value of the observed conversions determined in the one or more campaigns in the given traffic segment, according to the respective weight;

summing a weighted value of the expected sum of conversions of the one or more campaigns according to the respective weight, to thereby obtain a segment-specific expected conversions value; and

calculating the estimated quality score for the given traffic segment, based on a ratio between the third sum and the segment-specific expected conversions value.

14. The computerized device of claim 12 , wherein the one or more processors are further configured to perform at least one of (i) and (ii):

i. determine that an association exists between first and second campaigns of the plurality of campaigns; and

calculate an aggregated conversion rate of the first campaign using data that includes observed conversions and observed engagements of the second campaign,

ii. determine that an association exists between first and second traffic segments of the plurality of traffic segments; and

calculate a segment-specific estimated quality score of the first traffic segment using data that includes observed conversions and expected conversions of the second traffic segment.

15. The computerized device of claim 12 , wherein the one or more processors are further configured to:

for a specific campaign from the one or more campaigns, calculate an adaptive estimated quality score of the specific campaign indicating an estimated quality score of the specific campaign when presented in the given traffic segment, said calculate the adaptive estimated quality score comprising:

determining a fourth sum of observed engagements with the specific campaign when presented in the given traffic segment;

determining a fifth sum of observed conversions in the specific campaign when presented in the given traffic segment;

calculate a pair-specific quality score of the specific campaign based on a ratio between the fourth sum and the fifth sum;

calculate the adaptive estimated quality score of the specific campaign based on a combination of a weighted value of the estimated quality score of the given traffic segment and a weighted value of the pair-specific quality score.

16. The computerized device of claim 12 , wherein the one or more processors are further configured to generate an alert, in the event that the quality level is below the threshold.

17. The computerized device of claim 12 , wherein the one or more processors are further configured to, upon determining that the observed engagements comprise the at least some intent-less engagements or fraudulent engagements, perform one or more of:

generate an alert indicating a suspicion of an engagement fraud; and

execute a preventive action directed for mitigating the engagement fraudd.

18. The computerized device of claim 12 , wherein a reward is given per a user engagement with content of the one or more campaigns presented in the given traffic segment, the one or more processors are further configured to:

automatically update a value of the reward according to changes detected in the estimated quality score of the given traffic segment.

19. The computerized device of claim 12 , wherein the given traffic segment is associated to any one of: a website; a time frame; a group of users characterized by one or more common characteristics.

20. A non-transitory computer readable storage medium tangibly embodying a program of instructions that, when executed by a computer, cause the computer to perform a method

for each given campaign of a plurality of campaigns, obtaining information pertaining to engagements and conversions of the given campaign when presented in a plurality of traffic segments, wherein each of the plurality of traffic segments comprises traffic related to a plurality of users; and

calculating an aggregated conversion rate of each given campaign of the plurality of campaigns when presented in the plurality of traffic segments, said calculating the aggregated conversion rate of each given campaign comprising:

obtaining a first sum of observed conversions of the given campaign when presented in the plurality of traffic segments;

obtaining a second sum of observed engagements of the given campaign when presented in the first plurality of traffic segments; and

calculating the aggregated conversion rate of the given campaign based on the first sum and the second sum, wherein the aggregated conversion rate indicates a probability of a conversion to occur for an engagement with the given campaign when presented in any of the plurality of traffic segments; and

for a given traffic segment presenting one or more campaigns of the plurality of campaigns, calculating an estimated quality score of the given traffic segment, said calculating the estimated quality score of the given traffic segment comprising:

determining a third sum of observed conversions in the one or more campaigns when presented in the given traffic segment;

determining an expected sum of conversions of the one or more campaigns based on aggregated conversion rates of the one or more campaigns, as calculated in said calculating the aggregated conversion rate of each given campaign, and based on a number of observed engagements with each one of the one or more campaigns when presented in the given traffic segment; and

based on the third sum of observed conversions and the expected sum of conversions in the given traffic segment, calculating the estimated quality score of the given traffic segment, wherein the estimated quality score of the given traffic segment is indicative of a quality level of the observed engagements with the each one of the one or more campaigns when presented in the given traffic segment, wherein a quality level that is below a threshold indicates that the observed engagements comprise at least some intent-less engagements or fraudulent engagements in the given traffic segment.