R P Sundarraj

Smartphone applications are emerging as popular media for promoting one's products through in-app advertisements. Today, there are a number of organizations, known as supply-side-platforms (SSP), who aggregate and auction these ad-spaces from different suppliers/publishers. Advertisers (or their intermediaries) place bid for these spaces based on different relevance criteria (e.g., the location and device of the app-user, the app's IAB category etc.), the impression value, clickthrough value, and the conversion value. After the received ads are filtered based on relevance, the SSP is often still faced with a number of options for ad-placement, each having different revenues owing to differences in clickthrough rates etc. Moreover, the SSP has to decide on the ad-placement in real-time. In this paper, we consider the SSP's ad-placement problem in the aforementioned situation. We propose an optimization model to maximize the SSP's revenues. Based on computational experience with this model, we develop a rule-based online algorithm that appears to be viable as a real-time solution.

With the proliferation of new innovative technologies, electronic gadgets like smartphones loaded with sophisticated user friendly apps are getting popularity among large proportion of phone users. As a result, advertisers are using apps as a medium of communication to showcase their products through in-app advertising. These advertisements are managed by ad aggregators through DSPs & by ad-space aggregators through SSPs. We need to allow most profitable ads to be pushed into appropriate apps so that overall SSP's revenue is maximized.Inapp advertising helps both the SSPs as well as DSPs to earn revenue. In this paper, we aim to maximize the revenue of SSP by developing machine learning based rule inferring technique from the output of mathematical model. Our initial computational results show the efficacy of online algorithm over previous research.

Today, smartphone-based in-app advertisement forms a substantial portion of the online advertising market. In-app publishers go through ad-space aggregators known as Supply-Side Platforms (SSPs), who, in turn, act as intermediaries for ad-agency aggregators known as demand-side platforms. The SSPs face the twin issue of making ad placement decisions within an order of milliseconds, even though their revenue streams can be optimized only by a careful selection of ads that elicit appropriate user responses regarding impressions, clicks, and conversions. This article considers the SSP's perspective and presents an online algorithm that balances these two issues. Our experimental results indicate that the decision-making time generally ranges between 20 ms and 50 ms and accuracy from 1% to 10%. Further, we conduct statistical analysis comparing the theoretical complexity of the online algorithm with its empirical performance. Empirically, we observe that the time is directly proportional to the number of incoming ads and the number of online rules.

The cloud computing spot instance is one offering that vendors are leveraging to provide differentiated service to an expanding pay-per-use computing market. Spot instances have cost advantages, albeit at a trade-off of interruptions that can occur when the user's bid price falls below the spot price. The interruptions are often exacerbated since customers often require resources in bundles. For these reasons, customers might have to wait for a long time before their jobs are completed. In this paper, we propose a behavioral-economic model in the form of time-preference-based bids, wherein users are willing to use and bid for services at other times if the vendor cannot provide the resources at the preferred time. Given such bids, we consider the problem of provisioning for such service requests. We develop a time-preference-based optimization model. Since the optimization model is NP-Hard, we develop rule-based genetic algorithms. We have obtained very encouraging results with respect to standard commercial solver as a benchmark. In turn, our results provide evidence for the viability of our approach for online service-provisioning problems.