A building block of many models in empirical industrial organization is a characteristic space, where products are modeled as a bundle of characteristics over which consumers have preferences. The ability of such models to predict counterfactual outcomes depends on how well this characteristic space representation can capture substitution patterns. A limitation of existing methods is that product characteristics must be observable. In this paper, we extend a machine learning approach (Bayesian Personalized Ranking) that allows us to jointly learn latent product characteristics and consumer preferences from search data. We then show how this can be combined with existing demand estimation approaches to predict demand. Our application is to the hotel market, where we combine two datasets: consumers' web browsing histories, and hotel prices and occupancy rates. Using an event study design, we show that closeness in latent characteristic space predicts competition: hotels that are close to new entrants lose the most market share post-entry. We take a more structural approach to the 2016 merger of Marriott and Starwood, demonstrating that by using latent characteristics and consumer preferences learned from search data, we can substantially improve post-merger predictions of demand relative to standard baselines.

Current datacenters are based on server machines, whose mainboard and hardware components form the baseline, monolithic building block that the rest of the system software, middleware and application stack are built upon. This leads to the following limitations: (a) resource proportionality of a multi-tray system is bounded by the basic building block (mainboard), (b) resource allocation to processes or virtual machines (VMs) is bounded by the available resources within the boundary of the mainboard, leading to spare resource fragmentation and inefficiencies, and (c) upgrades must be applied to each and every server even when only a specific component needs to be upgraded. The dRedBox project (Disaggregated Recursive Datacentre-in-a-Box) addresses the above limitations, and proposes the next generation, low-power, across form-factor datacenters, departing from the paradigm of the mainboard-as-a-unit and enabling the creation of function-block-as-a-unit. Hardware-level disaggregation and software-defined wiring of resources is supported by a full-fledged Type-1 hypervisor that can execute commodity virtual machines, which communicate over a low-latency and high-throughput software-defined optical network. To evaluate its novel approach, dRedBox will demonstrate application execution in the domains of network functions virtualization, infrastructure analytics, and real-time video surveillance.

In this paper, we study the question and answer (Q&A) feature of electronic commerce platforms, an increasingly common form of user-generated content (UGC) that allows consumers to publicly ask product-specific questions and receive responses, either from the platform or from other customers. Using data from a major online retailer, we show that Q&As complement reviews and ratings: unlike reviews, Q&As primarily happen pre-purchase, focus on clarification of product attributes (rather than discussion of quality), and convey fit-specific information in a sentiment-free way. Our main hypothesis is that Q&As mitigate product fit uncertainty, leading to better matches between products and consumers, and therefore improved product ratings. We show that when low-rated products start receiving Q&As, their subsequent ratings improve by approximately 0.5 stars. We further show that the extent of the rating increase due to Q&As is moderated by the degree of ex-ante fit uncertainty. Overall, our findings suggest that, by resolving product fit uncertainty in an e-commerce setting, the addition of Q&As can be a viable way for retailers to improve ratings and sales of low-rated products, particularly those products that have incurred low ratings due to customer-product fit mismatch.

Two recent disruptions to the online advertising market are the widespread use of ad-blocking software and proposed restrictions on third-party tracking, trends that are driven largely by consumer concerns over privacy. Both primarily impact display advertising (as opposed to search and native social ads), and affect how retailers reach customers and how content producers earn revenue. It is, however, unclear what the consequences of these trends are. We investigate using anonymized web browsing histories of 14 million individuals, focusing on "retail sessions" in which users visit online sites that sell goods and services. We find that only 3% of retail sessions are initiated by display ads, a figure that is robust to permissive attribution rules and consistent across widely varying market segments. We further estimate the full distribution of how retail sessions are initiated, and find that search advertising is three times more important than display advertising to retailers, and search advertising is itself roughly three times less important than organic web search. Moving to content providers, we find that display ads are shown by 12% of websites, accounting for 32% of their page views; this reliance is concentrated in online publishing, e.g., news outlets) where the rate is 91%. While most consumption is either in the long-tail of websites that do not show ads, or sites like Facebook that show native, first-party ads, moderately sized web publishers account for a substantial fraction of consumption, and we argue that they will be most affected by changes in the display advertising market. Finally, we use estimates of ad rates to judge the feasibility of replacing lost ad revenue with a freemium or donation-based model.

Failure to meet a consumer's expectations can result in a negative review, which can have a lasting, damaging impact on a firm's reputation, and its ability to attract new customers. To mitigate the reputational harm of negative reviews many firms now publicly respond to them. How effective is this reputation management strategy in improving a firm's reputation? We empirically answer this question by exploiting a difference in managerial practice across two hotel review platforms, TripAdvisor and Expedia: while hotels regularly respond to their TripAdvisor reviews, they almost never do so on Expedia. Based on this observation, we use difference-in-differences to identify the causal impact of management responses on consumer ratings by comparing changes in the TripAdvisor ratings of a hotel following its decision to begin responding against a baseline of changes in the same hotel's Expedia ratings. We find that responding hotels, which account for 56% of hotels in our data, see an average increase of 0.12 stars in the TripAdvisor ratings they receive after they start responding. Moreover, we show that this increase in ratings does not arise from hotel quality investments. Instead, we find that the increase is consistent with a shift in reviewer selection: consumers with a poor experience become less likely to leave a negative review when hotels begin responding.

Spurred by technological advancement, a number of decentralized peer-to-peer markets, now colloquially known as the sharing economy, have emerged as alternative suppliers of goods and services traditionally provided by long-established industries. A central question surrounding the sharing economy regards its long-term impact: will peer-to-peer platforms materialize as viable mainstream alternatives to traditional providers, or will they languish as niche markets? In this paper, we study Airbnb, a sharing economy pioneer offering short-term accommodation. Combining data from Airbnb and the Texas hotel industry, we estimate the impact of Airbnb's entry into the Texas market on hotel room revenue, and study the market response of hotels. To identify Airbnb's causal impact on hotel room revenue, we use a difference-in-differences empirical strategy that exploits the significant spatiotemporal variation in the patterns of Airbnb adoption across citylevel markets. We estimate that each 10% increase in Airbnb supply results in a 0:37% decrease in monthly hotel room revenue. In Austin, where Airbnb supply is highest, the impact on hotel revenue exceeds 10%. We find that Airbnb's impact is non-uniformly distributed, with lower-priced hotels, and hotels not catering to business travel being the most affected segments. Finally, we find that affected hotels have responded by reducing prices, an impact that benefits all consumers, not just participants in the sharing economy. Our work provides empirical evidence that the sharing economy is making inroads by successfully competing with, and acquiring market share from, incumbent firms.

Daily deals sites such as Groupon offer deeply discounted goods and services to tens of millions of customers through geographically targeted daily e-mail marketing campaigns. In our prior work we observed that a negative side effect for merchants selling Groupons is that, on average, their Yelp ratings decline significantly. However, this previous work was primarily observational, rather than explanatory. In this work, we rigorously consider and evaluate various hypotheses about underlying consumer and merchant behavior in order to understand this phenomenon, which we dub the Groupon effect. We use statistical analysis and mathematical modeling, leveraging a dataset we collected spanning tens of thousands of daily deals and over 7 million Yelp reviews. We investigate hypotheses such as whether Groupon subscribers are more critical than their peers, whether Groupon users are experimenting with services and merchants outside their usual sphere, or whether some fraction of Groupon merchants provide significantly worse service to customers using Groupons. We suggest an additional novel hypothesis: reviews from Groupon users are lower on average because such reviews correspond to real, unbiased customers, while the body of reviews on Yelp contain some fraction of reviews from biased or even potentially fake sources. Although our focus is quite specific, our work provides broader insights into both consumer and merchant behavior within the daily deals marketplace.

Daily deal sites have become the latest Internet sensation, providing discounted offers to customers for restaurants, ticketed events, services, and other items. We begin by undertaking a study of the economics of daily deals on the web, based on a dataset we compiled by monitoring Groupon and LivingSocial sales in 20 large cities over several months. We use this dataset to characterize deal purchases; glean insights about operational strategies of these firms; and evaluate customers' sensitivity to factors such as price, deal scheduling, and limited inventory. We then marry our daily deals dataset with additional datasets we compiled from Facebook and Yelp users to study the interplay between social networks and daily deal sites. First, by studying user activity on Facebook while a deal is running, we provide evidence that daily deal sites benefit from significant word-of-mouth effects during sales events, consistent with results predicted by cascade models. Second, we consider the effects of daily deals on the longer-term reputation of merchants, based on their Yelp reviews before and after they run a daily deal. Our analysis shows that while the number of reviews increases significantly due to daily deals, average rating scores from reviewers who mention daily deals are 10% lower than scores of their peers on average.

Recently, some mainstream e-commerce web sites have begun using "pay-per-bid" auctions to sell items, from video games to bars of gold. In these auctions bidders incur a cost for placing each bid in addition to (or sometimes in lieu of) the winner's final purchase cost. Thus even when a winner's purchase cost is a small fraction of the item's intrinsic value, the auctioneer can still profit handsomely from the bid fees. Our work provides novel analyses for these auctions, based on both modeling and datasets derived from auctions at Swoopo.com, the leading pay-per-bid auction site. While previous modeling work predicts profit-free equilibria, we analyze the impact of information asymmetry broadly, as well as Swoopo features such as bidpacks and the Swoop It Now option specifically. We find that even small asymmetries across players (cheaper bids, better estimates of other players' intent, different valuations of items, committed players willing to play "chicken") can increase the auction duration significantly and thus skew the auctioneer's profif disproportionately. We discuss our findings in the context of a dataset of thousands of live auctions we observed on Swoopo, which enables us also to examine behavioral factors, such as the power of aggressive bidding. Ultimately, our findings show that even with fully rational players, if players overlook or are unaware any of these factors, the result is outsized profits for pay-per-bid auctioneers.

Attributing a dollar value to a keyword is an essential part of running any profitable search engine advertising campaign. When an advertiser has complete control over the interaction with and monetization of each user arriving on a given keyword, the value of that term can be accurately tracked. However, in many instances, the advertiser may monetize arrivals indirectly through one or more third parties. In such cases, it is typical for the third party to provide only coarse-grained reporting: rather than report each monetization event, users are aggregated into larger channels and the third party reports aggregate information such as total daily revenue for each channel. Examples of third parties that use channels include Amazon and Google AdSense.

This paper focuses on next generation ubiquitous networks supporting the Future Internet. In this context, it proposes an architecture and an integration framework of wireless and wired network technologies supporting a variety of services with differing service requirements. More specifically the i

This paper presents a programmable multi-granular optical cross connect (MG-OXC) and network architecture deployable in multi-service and multi-provider networks. The concept of programmable MG-OXC is introduced to provide a way of utilizing multiple switching/transport granularities to efficiently

This paper presents a novel solution for realization of service oriented optical networking. The solution is based on advanced optical burst switched network scenario utilizing novel service oriented optical burst switched core and edge router technologies. We demonstrate service-aware bandwidth reservation in a multi-granular OBS test-bed. We also demonstrate non-network services and data layer connections over OBS control plane by extending the JIT OBS protocol.

Grid-GMPLS (G2MPLS) is conceived as a powerful network control plane solution that enhances the standard ASON/GMPLS architecture providing single-step resource reservation, co-allocation and maintenance of both network and Grid resources. This paper identifies and discusses the main issues and considerations that arise by network research and educational networks and network operators in order to facilitate the dissemination of G2MPLS control plane. Interoperability issues and backwards compatibility with existing network control planes centre the scope of this study, which intends to demonstrate the feasibility of adopting the proposed architectures.

This paper presents a novel application-aware network architecture for evolving and emerging IT services and applications. It proposes and analyses network architectures that integrate Session Initiation Protocol (SIP) with Optical Burst Switched (OBS) protocols on a unified manner. We suggest various SIP-OBS layering architectures for possible deployment as well as a number of end-to-end resource discovery protocols (both for network and non-network resources). Finally the paper reports of a SIP-enpowered OBS Testbed where this approach was experimentally validated.

In many networked applications, independent caching agents cooperate by servicing each other's miss streams, without revealing the operational details of the caching mechanisms they employ. Inference of such details could be instrumental for many other processes. For example, it could be used for optimized forwarding (or routing) of one's own miss stream (or content) to available proxy caches, or for making cache-aware resource management decisions. In this paper, we introduce the cache inference problem (CIP) as that of inferring the characteristics of a caching agent, given the miss stream of that agent. While CIP is insolvable in its most general form, there are special cases of practical importance in which it is, including when the request stream follows an independent reference model (IRM) with generalized power-law (GPL) demand distribution. To that end, we design two basic "litmus" tests that are able to detect the LFU and LRU replacement policies, the effective size of the cache and of the object universe, and the skewness of the GPL demand for objects. Using extensive experiments under synthetic as well as real traces, we show that our methods infer such characteristics accurately and quite efficiently, and that they remain robust even when the IRM/GPL assumptions do not hold, and even when the underlying replacement policies are not "pure" LFU or LRU. We demonstrate the value of our inference framework by considering example applications.

This paper presents hybrid optical router architectures (both edge and core) to support user-defined bandwidth reservations for emerging and evolving applications over wavelength channels (circuit), optical bursts or even optical packets. The edge router is based on the deployment of application-aware IP packet classification and burst/packet aggregation algorithms as well as agile and intelligent optical resource allocation. The mechanism is responsible for per application switching (OCS/OBS/OPS) service selection and Differentiated Service (DiffServ) provisioning. The optical core router can support all the abovementioned switching technologies. Both are generic and able to support any type of current or future application.

This paper presents a novel hybrid optical burst/circuit switched (OBCS) ingress edge router solution towards ubiquitous photonic Grid networking. It is based on the deployment of Grid application-aware packet classification and burst aggregation algorithms as well as agile and intelligent optical resource allocation. The proposed solution utilises a generic and highly scalable multi-dimension classification mechanism able to provide wire-speed classification at high bit rates up to 40 Gbps. The mechanism is responsible for per application switching (OBS/OCS) service selection and Grid Differentiated Service (GridDiffServ) provisioning. A CoS-Traffic-Time-LEngth-Service-oriented aSembly (COST2LESS) algorithm has been proposed to smooth the incoming Grid traffic and provide some CoS differentiation and initial results are presented. An agile optical data transmission mechanism has been also implemented to map Grid traffic asynchronously into optical bursts or wavelength channels (data plane) based on user/application-specific requirements. Furthermore an optical burst Ethernet switched (OBES) transport mechanism has been implemented to transport out-of- band control plane signalling information.

This paper presents a novel solution towards ubiquitous photonic grid networking. It is based on the deployment of long-reached and high-bandwidth optical infrastructure while taking advantage of recent developments in optical networking technologies such as optical burst switching. The proposed solution utilises optical burst switching and active router technologies. It aims to provide a physical infrastructure able to fulfil grid application requirements and make efficient use of network resources.