Snap Way to Design Ad Ranking Service Using Deep Learning

Snap geometry recently posted a file Blog Posts on how they designed their ad categorization and targeting service using deep learning.

Showing ads to users is the trend of monetizing social networking platform. Snap’s ad rating system is designed to target the right user at the right time. Snap follows a few principles in designing such a complex system. They focus on providing an excellent user experience while maintaining user privacy and security. The following image shows the high-level structure of Snap Ads Ranking and Targeting.

Snap .’s Comprehensive Ad Classification Service Architecture

This system consists of some microservices. It contains a module called Ad Eligibility Filter. This module applies some filters such as checking ad budget, privacy, and ad policy compliance. Determines whether an advertisement is appropriate to display to the user. Ads from this module are passed to the candidate generation module. The filter generation module identifies relevant ads to users. Usually a machine learning model is very light Re-Call used in this unit. Once hundreds or a few thousand ads are identified, a heavy deep learning model is used to score the likelihood of the ads converting. In the end, the auction is run to select and score ads based on business rules, budget, etc. The feedback loop is used to collect more user interaction data to retrain models.

Ad rating systems face some unique challenges compared to general recommendation and search ranking systems. First of all, size, cost and latency. As mentioned in Blog Posts :

Our ML models work very broadly; Every day we make trillions of predictions using models trained on billions of examples. Such a large scale involves significant training and heuristics costs. These models also operate under strict latency restrictions. A combination of highly optimized training platform, inference platform, and model architectures is needed to keep cost and latency within acceptable limits.

Low prediction error rate is an important factor in choosing an advertisement. The error in predicting the outcome will propagate in the auction phase resulting in low quality data in the model’s training. The rapid change in advertising inventory is also one of the main challenges. This makes it difficult to get good embed Representation of advertisements.

For heavy ml models, in-service deep learning models are used as described in Blog Posts :

We make use of the latest multitasking models like MMoEs And the PLE To jointly predict multiple conversion events (eg, app installs, purchases, and subscriptions). Our models also use the latest high-level feature interaction layers such as DCN And the DCN version 2 as a building stone. We also optimize these models for inference cost and response time by splitting them into multiple towers, ie one for user features processing and one for advertising features. We illustrate these components and their interrelationships in the figure below.

A typical high-level structure used to classify Snap ads

Ad Ranking Systems is a use case recovery systems. Some cloud service providers such as AmazonAnd the The GoogleAnd the Microsoft It also offers these return systems as a service to its customers. There are some open source systems like Ad search systems On Github for further review.

.

Source

H3C Launches Channel Kickoff 2021 in Pakistan, Promoting a Win-Win Partner Ecosystem

H3C, a leader in digital solutions has initiated its second virtual event of the year in Pakistan — H3C Channel Kickoff 2021, Moving your business forward. The virtual event encourages global partners to “Move their business forward” by embracing new challenges and seizing opportunities alike, to jointly create more business value with H3C in the new year and beyond. In Pakistan, H3C has successfully certified more than 30 partners and over 60 engineers, and representatives of local partners and industry customers joined the virtual event.

“Since its official entry into the Pakistani market in 2019, H3C has been paving the way to support local customers in accelerating their digital transformation. We have made significant investments in the development of the local partner ecosystem, staff team and service system. Thanks to this, we have received increasing support from our partners,” said Davis Hui, the Head of H3C Global Channel Management.

“In addition to the latest products and solutions, H3C will continue to provide Pakistani customers with advanced ICT experiences building on our wealth of accumulated knowledge across various industries. H3C will also promote ecologically favorable policies, support local business partners to successfully complete digital transformation in the digital age, and leverage the power of digital to enhance inclusiveness.”

During the virtual event, H3C hosted the 2020 Partners Awards to recognize the important contributions of its Pakistan partners over the past year. Six of H3C’s Pakistani partners received the Top Sales Elite Award, Industrial Market Development Award, Solution Sales Elite Award, Top Service Sales and Delivery Award, and lastly, the Best Collaboration Award.

At the awarding ceremony, representative of the award winners said that, “It is a great honor to receive this award from H3C. H3C has a full portfolio of digital infrastructure products as well as a comprehensive one-stop digital platform in ICT field. H3C, relying on its high-quality products and professional team, has won the trust of an increasing number of customers across various industries here in Pakistan. We firmly believe that H3C will achieve greater success in the digital future, and we will also develop closer cooperation with H3C.”

H3C also shared its Digital Brain Project 2020 in Pakistan, demonstrating the leading capabilities of its “AI in ALL” intelligence strategy. The strategy aims to develop innovative smart applications with ecosystem partners and build digital brains for customers in their respective domains. It relies on Smart Digital Platforms composed of Digital Infrastructure, Business Empowerment Platforms, Active Security, and Unified O&M Services.

For the Pakistani market, H3C launched a Cloud-based management platform, Cloudnet, which will make it easier to manage H3C Wireless Lan (WLAN) devices, switches and routers. H3C also provides Blade server B16000, designed for fast-growing enterprises with the combination of powerful performance, added flexibility, unified management, and greater reduction of total cost of ownership (TCO). For small and medium-sized enterprises in Pakistan, H3C provides the latest generation version of UIS (Unified Infrastructure System) 7.0 with 4 new engines and 3 types of hardware, which combines almost all important elements in the data center, namely computing, storage, network, security, O&M and cloud, into a single box.

H3C Channel Kickoff 2021 is also scheduled to take place in additional markets such as Russia and Turkey, among others. The initiative embraces the power of digitalization with customers and partners worldwide, promoting the development of the entire ecosystem. In the future, H3C will further develop and expand its ICT ecosystem, including increasing support and capacity for overseas partners, working together to provide customers with more options, and contributing to the prosperity and development of the global digital economy.

About H3C

H3C is an industry leader in the provision of Digital Solutions and is committed to becoming the most trusted partner to the customers in their quest for business innovation and digital transformation. H3C offers a full portfolio of Digital Infrastructure products, spanning across compute, storage, networking, security and related domains, and provides a comprehensive one-stop digital platform that includes cloud computing, big data, interconnectivity, information security, new safety, Internet of Things (IoT), edge computing, artificial intelligence (AI) and 5G solutions, as well as end-to-end technical services.

For more information about H3C, please visit http://www.h3c.com/en/

Here’s how Big Data is revolutionizing modern Cricket

The Cricket World Cup is a complete feast for fans all over the world. Hundreds of thousands of Cricket enthusiasts tune in every day to watch live streams, re-watches, post-match analyses and some even keep themselves updated with live scores now readily accessible via the Internet. Over the years, Cricket and other major sports have undergone a massive evolution – both in terms of how they are watched and how they are played. Technology has not only made Cricket more accessible to fans around the world, it has also changed the entire dynamic of the sport.

Even up to a few years ago, Cricket was predominantly guided by an umpire. If that finger went up, the batsman would have abruptly walked off the field. The umpire had the final say in the cricketer’s fate in a match. Off the field, cricketers relied merely on physical practice to better their game. However, technical advancements, particularly in Data Analytics, have taken the sport to a completely different level.

Run rate, strike rates and averages have always been talking points for cricket. But now, with advanced data capturing techniques, players, coaches and critics can meticulously analyze every aspect of a cricketer’s performance. These insights can contribute significantly towards an enhanced performance and match strategy.

Likes of Hawk-Eye, Infrared HotSpot and Snick-o-Meter almost eliminates the need for an on-ground umpire. It tracks ball trajectory and provides a much more augmented viewing experience to the audience.

Lately, analytical advancement in the sports industry has turned its face towards improving the playing elements of the game. The bat sensor, introduced in 2017 by Intel, has taken a step towards revolutionizing cricket. Its prime goal is to generate data for every stroke that a batsman plays. The sensor is connected to a mobile app that collects data for back-lift angle, bat start angle, bat speed, power index and rotational angle of the wrists. The bat sensors were used by players in the last Champions Trophy and is seen in the current World Cup too. Australian batsman David Warner has been famously using a bat sensor in his practice sessions to further his batting technique and methodology.

With such advancements, players can receive instant feedback. Data collected by analysis tools, helps identify problems and provides players with the minor or major tweaks they need to take their performance up a notch.

Speaking of how data augments Cricket, Narayan Sundararajan, Principal Engineer & Director at the Intel New Technology Group explained,

“Cricket is a complex sport. It has twelve different strokes and each player has their own style, so there’s a huge number of variables. What we can do is coordinate metrics such as backlift angle and swing, then correlate that with Hawk-Eye data to show how effective the shot is and how it might be improved.”

“During the Champions Trophy match between India and Pakistan, Mohammad Amir was bowling brilliantly. So when India’s Virat Kohli came in to bat, he adjusted his bat backlift to cope with Amir’s 145mph deliveries, bringing it down to counter them. That’s the sort of subtle adjustment that teams are looking for. Using that data, we can come out with clear and actionable conclusions that wouldn’t otherwise be possible.”

While these advancements certainly help batsmen, bowling is one area which is still open and waiting for robust technology for real time analysis. Or is bowling action too complex to be analyzed by contemporary sensors? Surely, bowling is fifty percent of the game of cricket and it demands the progression- progression that could bring in quantification of different dynamics of bowling. So the question remains, will we be seeing technological development in this domain of cricket any time soon?

These questions are vital for the game. As all other international sports use data analytics, cricket on the other hand has a greater ability to leverage the vast amount of data generated. In the times to come, cricketing experts anticipate data analytics to play an important role in the way cricket is played and watched. The question remains, when?

Microsoft Forum Ilumina financial leaders in Karachi

The detailed sequence of events in the series, Microsoft, in collaboration with the Institute of Bankers Pakistan in Karachi finance executives a “Forum Big Data” arranged. explores new ways spectacular growth in the financial sector.

Support it and to provide the financial industry with groundbreaking technology initiatives, was a major Microsoft Pakistan. He helped to analyze specific industry scenarios where technologies like Big Data, etc. Insight HD, ML Azure, Azure Power Analytics, companies have made large profits.

Microsoft Pakistan lead the company in the country – Mr. Abid Zaidi said, “The Pakistan leaders in the corporate finance and other professional fields of powerful new technologies, data analytics business, which are essential for the competitiveness of the right decisions these professionals would like to thank. ”

During the sessions, industry and financial Pakistan were the heads of the heads of corporate finance; how mobile devices and social computing in an explosion in data and cloud connectivity to allow more innovation are causing.

Operations Director of the Institute of Bankers Pakistan (IBP) – Mr. Faisal Hussain declared, “In addition to focusing on technology, Microsoft is constantly training to companies worldwide. We will move forward and rapidly changing business environment to compete in the challenge to encourage us are fortunate to have a global technology leader, cleverly using the forefront of technology. as financial professionals, we plan, identify opportunities, controlling risks review and, finding new revenue models, sources of income to the use of large amounts of data have a clear view on the investment or portfolio management costs.

Speakers financial leaders, information technology (IT) are driving business results through smart use of many examples of how shared. Powerful new data, Microsoft Technologies, align their corporate strategy between management and IT professionals through collaborative efforts to nurture a culture of creative thinking and innovation in companies that are facilitated.