In 2006, Jeff Howe coined the concept of crowdsourcing. Crowdsourcing is an act of outsourcing a job, previously done by workers, to a large group of people in the form of an open call. Nowadays, crowdsourcing is used by many commercial and public organizations. A widely known example of an organization using crowdsourcing is Wikipedia Foundation Inc. It is a non-profit charitable organization maintaining an online encyclopedia called Wikipedia. Wikipedia allows any internet user to edit its articles. The users of Wikipedia are not paid for their contributions. Nevertheless, by July 2012, the English version of Wikipedia contained 3,988,490 articles. Several studies indicate that the quality of the articles in Wikipedia is similar to the quality of paid encyclopedias. On 15 of May 2013, Google stated that it would implement the crowdsourcing principle in Google Maps. The new Google Maps will (1) generate personalized maps for each user and (2) add in 3D tours of notable landmarks. These two features will be based on user-generated content. The new version of Google Maps allows the users to leave reviews of places they have visited (See Figure 1). Taking into account the popularity of Google Maps, such a review system will probably generate a huge number of review fraud cases. The reason is that fraudsters will be able to leave undeserved negative reviews (see Section 2) or posting their own positive reviews (see Section 3). Figure 1. A screenshot of the New Google Maps (source: Google, http://www.google.com/help/maps/helloworld/desktop/preview/)
- Leaving undeserved negative reviews Reviews have a crucial importance for the success of any business or organization. They reflect how customers are feeling about products, services, employees, and specific processes within the company. “The wheels of online commerce run on positive reviews” stated Bing Liu, a data-mining expert at the University of Illinois in Chicago. If potential customers read negative reviews from other customers, they will simply use the services of a competing business. Because the new Google Maps allows anyone to post a review of a place, companies willing to be in an advantageous market position may post undeserved reviews to their competitors. The new Google Maps may become a battlefield where companies fight by using the reviews as weapons. Such a war will not be in the interest of the consumers who will be often misled by the reviews. Companies willing to post negative reviews of their competitors in Google Maps may hire a freelancer from the global job marketplaces (e.g. Elance.com, oDesk.com, freelancer.com, etc.) Freelancers from India, for example, will charge an hourly fee of 10 USD for search engine optimization services, including posting negative feedback. Such a freelancer may post more than 100 negative reviews for an hour. Thus, in exchange for a payment of 200-300 USD, the freelancer will be able to post a huge amount of negative reviews concerning the competitors of the company that hired him. “It’s a flaw in the system that negativity can become so amplified. You can have a string of four and five star reviews, and then you get a string of one star reviews and it will torpedo your sales, because people will see those most recent reviews and it’s a warning sign to the potential readers. […] If there’s a reviewer that only leaves one star reviews, or if they’ve left nothing but a single negative review, they’re a carpet-bomber.”
- Posting one’s own positive reviews Posting one’s own positive reviews is another type of review fraud. By posting positive reviews concerning their own products, companies may attract many customers. While the Federal Trade Commission has issued Endorsement Guides in October 2009 stating that all online endorsements need to make clear when there is a financial relationship between the endorser and the advertiser, the Federal Trade Commission does not strictly enforce the Endorsement Guidelines. There are two main methods for leaving positive reviews, namely, using paid advertisers, or using automated tools. Below, these two methods are described in more detail. Paid advertisers A study by D. Mayzlin, Y. Dover, and J. Chevalier conducted in 2012 discovered that small hotels are likely faking reviews. The study indicated that such hotels are also producing fake negative reviews for nearby chain hotel competitors. These findings were based on an examination of half a million online hotel reviews published by Tripadvisor and Expedia.com. Tripadvisor is a website that collects and publishes customer reviews of attractions, hotels, restaurants, and other travel-related services. Any internet user can post a review on Tripadvisor. Expedia.com is a website that allows booking of airline tickets, car rentals, cruises, hotel reservations, vacation packages, and other attractions. It should be noted that most of the paid reviewers did not read the books or visit the places. They have to write a huge number of reviews a day, so they cannot spend a lot of time working on a single review. Automated tools Companies willing to gain a competitive advantage may use not only paid advertisers, but also automated tools for leaving reviews. XRumer is an example of a program that can be used for automatically generating positive reviews. XRumer is a search engine optimization program that can successfully register and publish posts to forums. The program has features allowing it to bypass security techniques used to prevent spam, such as account registration, CAPTCHAs, and post-registration email activation. CAPTCHAs (Completely Automated Public Turing test to tell Computers and Humans Apart) are used to prevent automated sign-ups to various types of online accounts. In order to demonstrate that the account is not registered by a computer, the user has to identify letters depicted in an image. This is an easy task for humans but tricky for computers. Once started, XRumer works fully automatically. The user only needs to add a text of the review, select a list of websites that he/she wishes to target, and press the “START” button. Crowdsourcing has proven to be an effective and efficient way to generate and maintain valuable information, such as customer reviews. However, crowdsourcing has one important disadvantage, namely, crowdsourcing systems are easy to manipulate. For example, in the case of the online review systems, anyone can post undeserved reviews or generate fake positive reviews. In the author’s opinion, Google implements crowdsourcing without taking sufficient safeguards against frauds, such as posting undeserved negative feedback. The introduction of the new Google Maps can give a rise to a new business of writing negative or positive reviews. We can call it “Google maps optimization.” Google could have learned from the experience of other systems using crowdsourcing, such as eBay and maarktplaats.nl. In order to solve the problem of undeserved negative feedback, eBay introduced the eBay’s Community Court. The Community Court allowed sellers to appeal non-positive feedback on matters that they believed to be unfair. Their claims were judged by 21 randomly selected members of eBay’s Community. The decisions of the jury were enforced by the eBay Customer Service Representative, who, if appropriate, removed the feedback. The eBay’s Community Court was capable of resolving a large number of disputes, free of charge, and in a short time. Moreover, the eBay’s Community Court prevented cheating by introducing the requirement that jury members would not be allocated to a case if they ever had a transaction with either side of appeal. On January 31, 2012, the ECRF was stopped. According to eBay, the reason for closing the ECRF was the reduced impact of negative feedback on a seller’s performance evaluation. The impact was reduced because eBay began evaluating seller performance based on low anonymous Detailed Seller Ratings (DSRs) which can rate specific aspects of the transaction. However, a Dutch online auction (www.marktplaats.nl) still offer a crowdsourcing court (http://gebruikersjury.marktplaats.nl) similar to the eBay’s Community Court. Crowdsourcing courts resolving disputes show that there are innovative solutions to the problem concerning negative user-generated content. These solutions need to be taken into account by Google when creating one of the biggest user-generated websites. Otherwise, the use of crowdsourcing in Google Maps may bring as much benefit as harm. References
