Recommender Systems in E-Commerce (1/2)

Recommender systems are tools for interacting with large and complex information spaces. They provide a personalized view of such spaces, prioritizing items likely to be of interest to the user. Recommender systems research has incorporated a wide variety of artificial intelligence techniques including machine learning, data mining, user modelling, case-based reasoning, and constraint satisfaction among others. Personalized recommendations play an important part in many online e-commerce applications such as Amazon.com, Netflix, and Google Search. This article provides an overview of how Recommender Systems are used in e-commerce, and the common types used in real-world projects. It consists of two parts: the first part will explain the theoretical knowledge, and it will use Amazon as an example. The second part will explain how to apply some of these techniques in python with a complete project on GitHub.

1. Introduction

The prototypical use case for a recommender system occurs regularly in e-commerce settings. A user, Mohamed, visits his favourite online electronics store. The homepage lists current a list containing recommended items.

This list might include, for example, a new product made by one of Mohamed’s favourite companies, a new video game. Whether Mohamed will find these suggestions useful or distracting is a function of how well they match his tastes. A key feature of a recommender system, therefore is that it provides a personalized view of the data, in this case, the electronics store’s inventory. If we take away the personalization, we are left with the list of best-selling products – a list that is independent of the user.

The recommender system aims to lower the user’s search effort by listing those items of the highest utility, those that Mohamed might be most likely to purchase. This, of course, is beneficial to Mohamed as well as the e-commerce store owner.

“If I have 2 million customers on the Web, I should have 2 million stores on the Web.”(Jeff Bezos, CEO of Amazon.com).”

2. What is a Recommender System?

There are so many definitions describe what Recommender Systems are, but I find this one the simplest: “any system that produces individualized recommendations as output or has the effect of guiding the user in a personalized way to interesting or useful objects in a large space of possible options.” (Burke, 2002)

3. Recommender Systems Techniques

There are various methods are being proposed to develop effective recommendation systems, two forms are the basis for the development of other approaches. These methods are content filtering and collaborative filtering.

4. Amazon.com

• 4.1 Information Page

many E-commerce sites are structured with an information page for each book, giving details of the text and purchase information. The Customers who bought feature is found on the information page for each book in their catalogue, and you will find two separate recommendation lists, and sometimes more.

The first section recommends other books frequently purchased with the selected book.

The second section presents books that are often visited by customers who have seen the same book.

• 4.2 Eyes

The Eyes feature allows customers to be notified via email of new items added to the Amazon.com catalog. Customers enter requests based upon author, title, subject, ISBN, or publication date information. Customers can use both simple and more complex Boolean-based criteria (AND/OR) for notification queries.

• 4.3 Amazon.com Delivers

Amazon.com Delivers is a variation on the Eyes feature. Customers select checkboxes to choose from a list of specific categories/genres (Oprah books, biographies, cooking). Periodically the editors at Amazon.com send email announcements to notify subscribers of their latest recommendations in the subscribed categories.

• 4.4 Book Matcher

The Book Matcher feature allows customers to give direct feedback about books they have read. Customers rate books they have read on a 5-point scale from “hated it” to “loved it.” After rating a sample of books, customers may request recommendations for books they might like. At that point a half dozen non-rated texts are presented which correlate with the user’s indicated tastes. Feedback to these recommendations is provided by a “rate these books” feature where customers can indicate a rating for one or more of the recommended books.

• 4.5 Customer Comments

The Customer Comments feature allows customers to receive text recommendations based on the opinions of other customers. Located on the information page for each book is a list of 1-5 star ratings and written comments provided by customers who have read the book in question and submitted a review. Customers have the option of incorporating these recommendations into their purchase decision.

5. Types of Recommender Systems

• 5.1 Non-Personalized Recommendations

Non-personalized recommender systems recommend products to customers based on what other customers have said about the products on average. The recommendations are independent of the customer.

Example: the average customer ratings displayed by Amazon.com and text comments supported in Amazon’s Customer Comments are non-personalized recommendations.

Advantages:

It’s useful when it comes to new users because we don’t have enough information about their preferences, so we can show them the bestsellers, popular author’s books..., until we form a picture of the user’s personality.

Disadvantages:

In the long term, it’s not useful since each customer gets the same recommendations, and not all customers have the same preferences, besides, their preferences change over time.

• 5.2 Attribute-Based Recommendations

Attribute-based recommender systems recommend products to customers based on the syntactic properties of the products. For instance, if the customer does a search for a historical romance book, and the E-commerce site responds with a list of three recommended books, that is an example of an attribute-based recommendation since it recommends the books that share the same properties of what they're looking for.

Attribute-based recommendations are often Manual, since the customer must directly request the recommendation by entering his desired syntactic product properties. Attribute-based recommendations can be either Ephemeral or Personal, depending on whether the E-commerce site remembers the attribute preferences for a customer.

Amazon.com Delivers 4.3 is Manual, since customers must explicitly sign up and provide a set of interest categories. However, Amazon.com Delivers is Persistent, since Amazon.com continues to send out recommendations in selected categories until the customer turns off the request.

• 5.3 Item-to-Item Correlation

It's similar to the previous type, but it's slightly different since the system executes a model-building stage by finding the similarity between all pairs of items automatically. For instance, if a customer has placed a few products in their shopping basket, the recommender system may recommend complementary products to increase the order size, a user enters the information page the system recommends products similar to they're browsing.

Ex: Back to 4.1, Amazon.com uses this type in its information page.

• 5.4 People-to-People Correlation

People-to-people correlation recommender systems recommend products to a customer based on the correlation between that customer and other customers who have purchased products from the E-commerce site.

The system tries to detect the patterns among users and based on the correlation it tries to recommend products. For instance, the rate of each user for a book if the rate is positive, and let’s say that a user gave a nice comment 4.5 about this book, this makes the book more apperant on the main page. because the system discovered that many people like this product.

• 5.5 User Inputs

Each of the previous four recommendation technologies requires some form of input upon which to base the recommendations. Typically this input is provided by the customer(s). However, it is possible that the input may also be provided by the business as well. The systems in our examples utilize one or more of the following inputs.

Purchase data: Which products a customer has purchased. Systems such as Amazon.com’s Customers who bought a certain book make recommendations based entirely patterns of “copurchase” between multiple customers.Get back to 5.4.

Likert: What a customer says he thinks of a product, typically on a 1-5 or 1-7 scale. The scale may be numeric or textual, but must be totally ordered. Systems such as Amazon.com’s Customer Comments.Get back to 4.5 and, 5.4.

Editor’s choice: Sometimes selections within a certain category are made by human editors, usually employed by the E-commerce site.

6. Conclusion

The importance of recommendation systems is important when it comes to creating an individual experience that suits the interests of each individual. In fact, we are not only witnessing it in e-commerce, but in almost everything and without it, the user experience will be pretty difficult. The problem sometimes lies in the unethical use of individual data, so there must be regulations regulating the use of each person's data.