Introduction
This report discusses Netflix recommender system by examining how the organization use data science to manage the massive data it holds to serve its users. Recommender systems are common in e-commerce organizations, content-based websites. It involves a technology used to analyze data in order to make suggestions for what may interest users. Netflix is one of the organizations that use recommender system to recommend videos to its users.
What is Netflix?
Netflix is one of the most popular online subscription platform offering television shows and movies to its customers through mail or by direct streaming from their websites (Netflix, 2019). This is provided through streaming service where one can watch videos as such, movies, television shows well as documentaries by subscribing and streaming using data to access the internet.
Relation to data science
The organization initially started as a DVD retail service provider back in the year 1998. For its business, they majorly relied on the third party postal services to serve their customers which did not show some efficiency which the organization mitigated soon with the use of data science. (Muvi, 2019). Netflix apply data science in various ways, however, for this report, we focus on recommender system.
Movie recommendation system:
The recommendation system easily identify users’ needs and offer suggestions of various video products to the Netflix users. This is achieved by use of watching history of their customers with alike tastes for recommending what their customers would be most interested in watching next time so that they engage them and continue subscribing for more shows.
The recommender system employs various technology and methodologies in filtering massive data and provide a smaller amount of the data for users to suggest the ones that fulfill their interest. In Netflix, Meta data tagging is used in videos along with the data about the behavior of Netflix users in order to come up with recommended videos as well as TV shows for specific users.
Data Roles
By looking at the Netflix official website, we can conclude that there are many data roles practiced in Netflix Company.
Data Analytics: Involves an analyst collecting and storing data and information of behaviors (Waller & Fawcett, 2013). They have expertise that ensure data accuracy and process it as well presenting it in a way that can be meaningful.
Data Algorithms: It involves data mining and use of data analysis methods. They build and implement models as well as creating simulations.
Data analyst
Netflix Company’s data analyst is an individual who analyses data with an aim achieving some insights. These insights are then transmitted to the management of the company for making of decisions. Therefore, as a data analyst at this organization a person should be able to examine data at all levels be it those from reports or databases.
Machine learning engineer
This is a person that owns a degree in machine learning or related field. This person should be able to use skills of machine learning to a widespread dataset. Also, he/she should come up with solutions that can create a strong user base of Netflix Company.
Business Model
With much revolution in the industry, there is rise in number of internet users who search for streaming services. It is important to note that, even such services came to existence; people opted for CDs and bought films and videos from video shops. Netflix utilized the rate of mobile acquisition and use of internet to provide the services. Thus, they provided the internet users with relevant and their desired videos, films, documentaries and TV shows. Netflix’s success solely depends on some of the aspects namely: the capability to ensure updated categories of videos through analysis of users’ demand and existing up-time search engine optimization (SEO).
The organization makes use of a subscription based business model. Initially, they have been using a business model which is quite similar to the brick and mortar video rental store. Netflix is exploring the change in the way it provides services to its consumers from DVDs to online streaming and it is now turning to a global television network. The organization buy products from product creators and distributes the products to its subscribers through internet
Characterising The Data And Data Processing
Data types
Netflix has various data in its store. The following shows the data types in the organization. The data types are divided into two, firstly the actual films and movies types that users want to view. This they can stream live or download on their devices.
Videos and movies: these are product files supplied to the organization’s subscribers
Customer feedback and interaction data: these include user response, time and dates users watch as well as the device used.
Netflix Company has kept in their database more than eight thousand TV shows and movies. It has 2 types of data storage; Flash drive and a hundred terabyte hard drive (McCord, 2014). The main challenge that they encountered is coming up with a readily available storage technique that would keep their data for long. Cloud storage was adopted as its remedy.
Reaching customers
Netflix Company allows their users to subscribe so that they can access the content of choice that have been arranged in libraries.
Processing times
Netflix company process data in two ways; periodic and real time. This data processing is achieved by Aesthetic Visual Analysis (AVA) technology.
Data storage and volume
Netflix organization have storage servers made of a combination of hard drives which are combined together in a server. The organization currently uses 36 drivers which holds approximately 100 tetra bites of data. The servers can store and stream approximately 15,000 movies simultaneously and they have almost thousands of data stations spread around the world where contents get collected after which they are spread to various places around the world. The volume of data held by Netflix requires to be organized and shifted. Thus during the off hours, the organization always fill the servers with movies and shows which are most popular in order to reduce the band width during high seasons.
Netflix data handling technology layer is called Extra Transform and Load ETL. This approach include the process of managing big data where methodologies which take into account features including quality are used. By this, the data which are useful to the organization are used as it is rarely associated with errors and it corresponds to the business goals of the organization.
Data loading process
The data loading process is outlined within the data plan. It outlays facets like data storage and data management models which are always available in two options including: cloud storage and big data repository i.e. data warehouse.
Netflix data analytical layer
This is the process that the organization executes once the data is ready i.e. once the massive data that is needed by the organization is ready. The process provides the previously established performance metrics. They have a catalog which relies on complex algorithms for monitoring its subscribers online which also aims at formalizing the satisfactory decisions regarding views.