Here's Why We Built Open Source Infrastructure for Open Reviews


We built a service for online reviews that makes the insights you share accessible across any application, and that does not invade your privacy. Leave proprietary data silos behind and enter the open data ecosystem of Mangrove reviews. 


Our goal for this blog post is to show you around in the newly released Open Reviews web app: this is a brief summary of its capabilities at a glance. 


Before we dive in, let's have a quick look at the bigger picture 

The insights that people share with each other via online word-of-mouth are an important, if not the only, way for us to discover, evaluate, and select the experiences that best fit our needs. Things become less risky for us to try ourselves, if someone else has already tried them before and has shared an opinion about them.

The world is becoming increasingly interconnected, with an unprecedented amount of offerings on a global scale. This creates complexity, and oftentimes even a paralysis of choice. But with tools that make discoverability easier, reduce risk, and augment consumer voices, this rich offering can lead to better experiences for the individual, and drive targeted improvement of products and services.

Word-of-mouth is a fundamental, age-old form of human collaboration. To make use of it on a global scale and to reap its potential to help us navigate the world, we need to apply technology smarter than we do today.

In brief, we are working on tools that enable the following:

  1. Stop adding reviews into data silos where the insights are partitioned and the user bases are divided
  2. Stop making these valuable crowd-sourced insights proprietary and only usable by a handful of dominant platforms
  3. Allow entrepreneurs, researchers, and others to experiment with this data, and to develop new ways to derive insights from it

Stay tuned for upcoming blog posts where we share more about our vision for creating an open data ecosystem for reviews. 

So, what can the Open Reviews web app do for you? In brief, it lets you read and write reviews! But what’s different to the well-known proprietary services out there? 

  • You can read/write reviews for many more things that matter to people, all in one place. These things are currently: places on the map such as restaurants, hotels, touristic sites; but also companies, websites, and books. You can share with people a wider range of your experiences if you like, all in one profile. 
  • The valuable insight you share with others is under a free license where it can be useful for many more purposes than proprietary services allow. On proprietary platforms, you can only  'consume' the data. In an open dataset it can be a rich source for research, open innovation and entrepreneurial activity.
  • Your contribution can never be sold off or shut down, because it is not an asset of one corporation, but a free cultural work. Adding your insight to an open dataset means it can and will be integrated into a variety of applications and websites. If any of them shuts down, there will be many left that mirror the same data and make your contribution still accessible.
  • Read/write reviews without being exploited for your personal information. We are strong advocates for privacy, and we developed a way to create accounts without requiring registration. You don’t need to entrust yet another service with your name, email, birth date, address, etc.. The Open Reviews web app and website don’t have any trackers, third-party cookies, or ads. You share any information consciously and voluntarily, directly with the public open dataset.

Now, let's explore it! 

Start your tour of the Open Reviews web app by entering something into the search box. This can be the restaurant where you ate last night, or a website you love and really want to recommend ...or a book worth sharing your thoughts about.

Search


Let’s try “Bürgenstock”, a picturesque village and touristic spot on a hill above Lake Lucerne that we’ve recently visited.

Press enter, and the Results page appears. On the left you will see a list of subjects from different categories that are related to your search word. For “Bürgenstock” there are places, companies and books coming up. You can use the category filters to display only the relevant results, in this case “Places”.

Once you click on the subject of your interest, you will see on the right its details and all reviews that have been submitted for it. If it’s a place, you will also see its location on the map. You can use the map to navigate to a different place and select that instead.


Each of the categories that can be reviewed draw their identifiers from other open datasets. Places for instance use OpenStreetMap to identify a point of interest or geographic location. For books we use ISBN via the Open Library. For companies we use the Legal Entity Identifier provided by GLEIF. Websites are identified by their URL.

Now, let’s share some experiences and insights - where if not here, in the public, where it belongs to humanity as a whole and is not part of just one corporate’s data treasure trove that can be sold off or shut down at the whim of its owners.

Writing a review

Clicking the “Write a review” button opens the review form. It is slightly different for each category. Let’s look at one for books. What should be familiar from proprietary services are the stars, the text box to describe your experience, as well as the “I agree to the Terms of Service” confirmation. 


What is, however, significantly different is that here you are in full control of what information you share about yourself. 

Privacy

In today’s dominant proprietary review services you share one type of information openly on the platform for other users to see. That’s the voluntary part. But you share a whole lot more information in the background. 

While being logged in on the platform, you are exposed to trackers, and first and third-party cookies that record your search and browsing history, as well as your online behavior to determine predictive patterns that make you an easier target for advertising and political campaigns. Your data is shared across a plethora of intertwined parties with contractual relationships that are practically impossible to dissect. 

Harvard Professor Shoshana Zuboff, the author of the best-selling book 'The Age of Surveillance Capitalism’, has called this a 'form of tyranny that feeds on people but is not of the people'.

Our approach is different. In the Open Reviews web app, there is just one type of information shared by users: the one entered consciously and voluntarily in the review form. Once submitted, it goes directly into the open dataset, with no intermediary collecting anything in the background.

The dataset is open for everyone to inspect, and the whole code base, comprising the standard, the server, the user interface, as well as the APIs and the JS client library, are open source and allow for scrutiny by the community to ensure that it does what it promises.

Account, redefined

In a reviews system, identity and reputation play an important role in ensuring reliable ratings. To allow for persistent identity while ensuring high privacy standards we introduced a new way for users to create their identity. It does not require any central party to keep track of it.

Users interact directly with the database. 

Open Reviews leverage public-key cryptography to enable that. The user's device generates a secure password (a cryptographic private key), allowing them to write reviews under a unique identifier (symbolized by an avatar), and to authenticate themselves as its rightful owner. 

This mechanism does not require any intermediaries. Simply saving the private key, ideally in a password manager, and re-entering it in case the browser’s local storage was erased, allows users to access their account at all times, from any device.



The aggregation algorithm that computes ratings based on individual reviews gives higher weight to reviewers with a good reputation, i.e., a track record of high-quality reviews. 

Reputation is built by a number of things, such as contributing regularly reviews that display a differentiated assessment, by other users marking reviews as “useful” or “confirming” the reviews as matching with their own experience, or by providing proofs of purchase (e.g., pictures, scans of bills). 

Reputation is diminished in cases where users flag a review as inappropriate, or if a reviewer displays behaviour that violates the Terms of Service.

Reviews can also be commented on or replied to by other users. This is particularly interesting for owners of businesses or websites interacting with their customers, as well as authors replying to readers.

Build your profile 

Are you a passionate traveler and would like to share your recommendations and experiences freely with others, outside of proprietary data silos? Mangrove is made for that! Not only can the insights you share be integrated into your website or app, they can also be leveraged freely by travel & tourism blogs, online magazines, and more - so that many more people can read and benefit from your experiences. In Mangrove, you can enrich your recommendations for places with reviews for books, websites, and more, to create a more rounded view on what matters to you. 

Check out this new user's travel recommendations to Argentina and Uruguay


Integrate Open Reviews into your app or website

This open-source infrastructure is built for integration. Ultimately, we’ve got to build successful ecosystems around open data to get to the scale at which we can speak of meaningful alternatives to proprietary systems. 

Check out how you can integrate Mangrove into your application or website, and stay tuned for a dedicated blog post about this topic.

Join the Open Reviews Association

This is a non-profit service hosted and supported by the Open Reviews Association, a community of businesses, developers and interested parties around the world. You can join the association and participate actively in the decision-making and further development of the project.  

If you are using apps or websites that could benefit from a reviews layer (any navigation map alternative to Google Maps, tourism websites, online libraries, and many more), please get in touch with us or with them directly to let them know. 

Comments

Popular posts from this blog

Gästebewertungen im Tourismus: Warum sie allen gehören sollten