Rakuten’s Journey to HTTPS – Lessons So Far
Rakuten group is currently migrating its services to HTTPS; any service that uses HTTP is being migrated to HTTPS. This includes not just websites but also our mobile applications and other services we offer, for example ad and content services. This work is still ongoing, however globally over 80% of Rakuten’s services had completed their migration by March 2017. We wanted to share our lessons for managing such a large scale migration.
Hello, World, this is Yu-Lu “Chris” Liu, I am the Office Manager of the IT Security Engineering Office, the technical security team at Rakuten, Inc. (hereby Rakuten). I am also a member of Rakuten-CERT. I would like to share about our on-going activities with universities. We believe we are nourishing the next generation of security pioneers, and contributing to society while protecting Rakuten’s users.
In recent times, as our reliance on technology increases, so does the threat of online cybercrimes.Online Criminals are dedicated to stealing your online personal information and identification. Important details like user IDs, passwords, credit card or bank information as well as personal information are all at risk of theft from cybercriminals. These online thieves utilize phishing emails, or fraudulent websites as their methods of attack. Whether through email or the web, cybercriminals are sophisticated in making their attacks appear as if they are legitimate websites or messages. The even utilize phishing email schemes, to appear as though they are sent from Rakuten. One real-world example of this is an email that appeared to be sent from Rakuten with the email subject “Congratulations”. These cybercriminals attempted to lure the user with a promise of a prize by inputting their personal information into a fraudulent website. These falsified websites are artistically deceitful as they are often meticulously modeled after Rakuten, utilizing Rakuten’s logo, merchandise photos and descriptions taken directly from one of Rakuten’s services. These websites even feature an e-commerce shopping basket, helping to craft the illusion they are a legitimate service. Of course, no order will ever actually ship. This occurrence becoming more and more common has made all users question the safety and security of online services.
If you’re an old-fashioned romantic, there’s no better time to express your love for that special someone than Valentine’s Day. Some Romeos go beyond the traditional chocolates and roses and offer their Juliets an engagement ring. But even if you don’t pop the question on February 14, all the work involved in planning a wedding can take away from the high of getting hitched. Researchers at Rakuten Institute Technology, though, are taking some of the pain out of the process with a platform that allows users virtual previews of possible venues for the big day.
In relation to deep learning, the two-layer network Word2Vec developed by Google was revolutionary. If you input text, you can obtain a vector set as the output, i.e., the feature vector of words in the text. By grouping the words and creating vectors, you can make judgments on similarity, and infer the meaning. Since it uses figures, it is scalable by processing in parallel and this is one of the great advantages of Word2Vec. It is not only applicable to the field of text analysis, but to any kind of data. At Rakuten, we have developed an extended version called Category2Vec and have released it as OSS (ref: https://github.com/rakuten-nlp/category2vec). We are training it with various kinds of EC data and are beginning to see the possibility of application on a broad scale for product/user analysis and categorization, and estimation of loss data. While Word2Vec is not deep learning itself, it is worth understanding this as a technology in numerical form that can be used in a deep neural network.
When studying machine learning, you will sometimes come across the term “online learning”. For example, in supervised learning such as SVM, first you train with all of the sample data given. Budepending on the volume of sample data or the application, it may not be appropriate to learn all the data at once. As mentioned above, all businesses have time restrictions. There are also system limitations on the volume of data that can be handled, such as the computer processing capacity and memory capacity. There are also cases where the sample data is provided in stages. In these cases, it can be convenient to optimize the parameters for each data set provided, and then revise it and retrain. This kind of method is called online learning. You could also call it “consecutive leaning”. Typical online learning methods are perceptron, CW, AROW and SCW. There is also a method called Streaming Random Forests, which is Random Forest by online learning.
Omotenashi (Hospitality) & AI
This blog is based on my speech at IBM Watson Summit on April 28th, 2017 following to joint press release of Rakuten AI Platform by Rakuten and IBM. ( https://global.rakuten.com/corp/news/press/2017/0426_01.html)
The Rakuten AI Platform has intended to facilitate various AI applications such as automated chat bot for customer services to be utilized in wide spectrum of Rakuten services including E-commerce, FinTech, Digital content and etc.
On the other hand, unsupervised learning is a method in which actual data is analyzed without receiving any prior sample data so that the intrinsic structure and characteristics existing in the data are extracted. For example, in recommendation, which is an important function in EC, we often use data clustering, which is an unsupervised learning method that categorizes the customers and products to recommend. As well as supervised learning, unsupervised learning is also used for security purposes. When detecting log-in attacks, we deploy clustering to determine what kind of attack patterns there are.