投稿時間:2021-12-14 06:22:50 RSSフィード2021-12-14 06:00 分まとめ(30件)

カテゴリー等	サイト名等	記事タイトル・トレンドワード等	リンクURL	頻出ワード・要約等/検索ボリューム	登録日
TECH	Engadget Japanese	2005年12月14日、通信・通話・アプリ利用がこれ1台で可能になった「W-ZERO3」が発売されました：今日は何の日？	https://japanese.engadget.com/today-203051301.html	windowsmobile	2021-12-13 20:30:51
AWS	AWS Architecture Blog	Use Amazon EKS and Argo Rollouts for Progressive Delivery	https://aws.amazon.com/blogs/architecture/use-amazon-eks-and-argo-rollouts-for-progressive-delivery/	Use Amazon EKS and Argo Rollouts for Progressive DeliveryA common hurdle to DevOps strategies is the manual testing sign off and deployment steps required to deliver new or enhanced feature sets If an application is updated frequently these actions can be time consuming and error prone You can address these challenges by incorporating progressive delivery concepts along with the Amazon Elastic Kubernetes Service Amazon EKS …	2021-12-13 20:49:01
AWS	AWS - Webinar Channel	AWS WEBINARS - Introducing data tiering for Amazon ElastiCache for Redis Afternoon Session	https://www.youtube.com/watch?v=QxIxhjfTtD4	AWS WEBINARS Introducing data tiering for Amazon ElastiCache for Redis Afternoon SessionDevelopers use Amazon ElastiCache as an in memory cache and data store to boost application performance and reduce latency to microseconds Today we re announcing data tiering for ElastiCache for Redis so you can scale and store data more cost effectively saving up to per GB Data tiering transparently moves data between RAM and solid state drives SSD and automatically optimizes the cost of your ElastiCache clusters with minimal performance impact Join us to learn how to get started with data tiering for ElastiCache and hear how Rokt a global leader in ecommerce marketing technology uses data tiering to simplify scaling and reduce costs	2021-12-13 20:59:19
Program	[全てのタグ]の新着質問一覧｜teratail（テラテイル）	keras-yolo3、文字コードに関するエラーを解決したい	https://teratail.com/questions/373679?rss=all	vocannotationpyを実行すると以下に示したエラーメッセージが表示されました。	2021-12-14 05:14:25
海外TECH	Ars Technica	Microsoft will sell you a $25 poster to celebrate a major Xbox hardware failure	https://arstechnica.com/?p=1820226	failureit	2021-12-13 20:15:25
海外TECH	MakeUseOf	How to Set Up any Email (Including Gmail and Exchange) on Mozilla Thunderbird	https://www.makeuseof.com/how-to-set-up-email-gmail-exchange-on-mozilla-thunderbird/	How to Set Up any Email Including Gmail and Exchange on Mozilla ThunderbirdAny type of email account can be set up in Mozilla Thunderbird POP IMAP Gmail or Microsoft Exchange here s how to get your email on Thunderbird	2021-12-13 20:30:12
海外TECH	MakeUseOf	Sony Launches New PlayStation 5 Covers: How to Pre-Order One Today	https://www.makeuseof.com/sony-launches-new-ps5-covers-how-to-pre-order/	Sony Launches New PlayStation Covers How to Pre Order One TodayBored with your PS s standard design You re in luck Sony has finally revealed new PS console covers and more DualSense controller colors	2021-12-13 20:07:58
海外TECH	MakeUseOf	75+ Adobe After Effects Keyboard Shortcuts to Make Your Life Easier	https://www.makeuseof.com/adobe-after-effects-keyboard-shortcuts/	Adobe After Effects Keyboard Shortcuts to Make Your Life EasierGet familiar with these Adobe After Effects shortcuts to instantly access various tools with your keyboard to save on time and effort	2021-12-13 20:02:52
海外TECH	DEV Community	My list of useful git commands	https://dev.to/tkarropoulos/my-list-of-useful-git-commands-5f9g	My list of useful git commandsAs all those of us involved in the field of Computer Science know or should know Git is by far the most widely used modern version control system in the world today We utilized it in our everyday routine to keep track of our code changes and helps us to work with other developers simultaneously and independently Although modern IDE and various tools provided by GitHub Atlassian and other provide us an easy to use way to perform many git commands nothing compares to the power a terminal provides Bellow you can find a list of my top git commands Rename latest s commit message This will pop up an editor window allowing us to pass the new commit messagegit commit amend This will not pop up the editorgit commit amend m Your new commit message Add file s into the latest commit This requires that last change is not yet pushed into remote Add the filegit add the file you want to add Amend without changing commit messagegit commit amend no edit Reset to specific commit hash and discard any changes since that hashgit reset hard lt commit hash gt Apply a commit from one branch to anothergit cherry pick lt commit hash gt Show commit logs and limit the outputgit log n lt number gt example git log n orgit log lt number gt example git log If you are aware of any useful useful git command and want to share it please do not hesitate to leave a comment	2021-12-13 20:47:27
海外TECH	DEV Community	Data Science: The Best Programming Languages for Data Scientists	https://dev.to/imagescv/data-science-the-best-programming-languages-for-data-scientists-42pm	Data Science The Best Programming Languages for Data ScientistsData Science is a growing field that has made it possible to extract meaningful information from data Data scientists need programming languages in order to have the necessary tools for their job In this blog post we will talk about the best programming languages for data science and why they are so important Programming languages used for data science R LanguageR Language a language and environment for statistical computing It is widely applied in research academia industry and business R programming has been around since but really got popular after when it was adopted by Kaggle which hosts machine learning competitions Many of the top places on Kaggle are occupied by R users PythonPython is a general purpose high level programming language with dynamic semantics and has been around since Python was designed to be highly readable which makes it an appropriate tool for data science due to the readability of code that can be used in machine learning algorithms or statistical procedures such as linear and non linear regressions C C the C programming language was created in and over time it evolved into a superset known as C however its syntax remains similar due to staying dedicated for systems level processing which makes it popular with those who work on operating system kernels the central component such as Linux C has a focus on performance and is used in applications which need to process large amounts of data or require the execution of computationally intensive tasks such as machine learning algorithms JavaJava was created by Sun Microsystems Inc now Oracle Corporation in and it runs on an environment known as Java Virtual Machine JVM Java is a versatile language that can be used for developing applications across various platforms Java has become popular with data scientists due to the advent of big data and its ability to handle large datasets Additionally there are many libraries available in Java for data science tasks such as machine learning or scientific computing Wolfram LanguageWolfram Language was created by Stephen Wolfram and is a highly technical language that has been gaining traction in the academic community It was first used to power Mathematica which can be considered as an environment for computation visualization etc The Wolfram Language is being adopted by some data scientists due to its ability to execute symbolic computations very fast along with having high performance libraries for data science tasks Which programming languages are used for data science R Python SQL C Java and Wolfram Language Each of these languages has unique benefits that make them popular among data scientists For example R is widely adopted due to its ability to be easily understood by those who are not experts in the field and Python is popular due to its readability and a large number of libraries C has become very common in the field since it was designed for systems level processing which makes it a great fit for those working on operating system kernels such as Linux Java became extremely popular with data scientists after big data came into existence along with having many useful libraries available for data science tasks and Wolfram Language is being adopted by some due to its symbolic computation abilities and high performance libraries Each of these languages has something unique to offer making them a great choice for those working in the field of data science Data science is a high demand field and knowing which programming languages data scientists should know can make or break their careers If you want to be a successful data scientist it s important that you have the right skillset We recommend learning Python for beginners as well as Java for more advanced work with computer code These two languages are used by over half of all Fortune companies so they re worth getting acquainted with images cv provide you with an easy way to build image datasets K categories to choose fromConsistent folders structure for easy parsingAdvanced tools for dataset pre processing image format data split image size and data augmentation Visit images cv to learn more	2021-12-13 20:20:31
Apple	AppleInsider - Frontpage News	Apple University dean Joel Podolny exits Apple for startup	https://appleinsider.com/articles/21/12/13/apple-university-dean-joel-podolny-exits-apple-for-startup?utm_medium=rss	Apple University dean Joel Podolny exits Apple for startupThe head of management training at Apple Joel Podolny reportedly left the iPhone maker earlier in departing the organization to join a startup company The dean of the Apple University training school Podolny allegedly left the company earlier in the year in favor of starting afresh with a new startup It is unclear exactly when he left nor the identity of the startup he joined but people familiar with the situation allegedly confirmed the departure A representative for Apple declined to comment about Podolny s apparent exit to Bloomberg Sources claim that he has been replaced by two new co deans of Apple University who previously reported to him but now report to Apple chief of retail and human resources Deirdre O Brien Read more	2021-12-13 20:54:52
Apple	AppleInsider - Frontpage News	How to: 4 ways to get into a locked iPhone without the Password	https://appleinsider.com/articles/21/12/13/how-to-4-ways-to-get-into-a-locked-iphone-without-the-password?utm_medium=rss	How to ways to get into a locked iPhone without the PasswordApple devices are popular for their security features including a strong passcode lock preventing unauthorized access But if you forget the passcode you ll have a problem Here s how to bypass it Locked iPhone No problem The passcode lock is an effective security mechanism for iPhone devices Put simply you won t be able to access your device if you can t remember the passcode And even worse if you have entered the wrong passcode or more times your iPhone will be disabled Read more	2021-12-13 20:23:10
海外TECH	Engadget	Apple delays macOS Universal Control until spring 2022	https://www.engadget.com/apple-universal-control-feature-mac-ipad-delayed-202744636.html?src=rss	Apple delays macOS Universal Control until spring All the way back at WWDC Apple showed off a feature called Universal Control which will let users control multiple Macs and iPads with a single mouse and keyboard or trackpad When it released macOS Monterey in October Apple said that feature and SharePlay would arrive on Macs later in the fall Although SharePlay is now available on Mac you ll need to wait a bit longer for Universal Control Apple quietly updated its website to state that Universal Control won t be available until spring The delay which was spotted by to Mac might come as a disappointment to those who were hoping for more seamless connectivity between their devices in the near future Still it s better to make sure the feature is working correctly instead of releasing a potentially buggy version When it does arrive the feature will be available on MacBook and Macbook Pro and later the Mac Pro MacBook Air and later iMac and later and the K Retina inch iMac from late As for supported tablets you ll be able to use iPad Pro iPad Air rd generation and later iPad th generation and later and iPad mini th generation and later with Universal Control You ll need to be logged into iCloud with the same Apple ID on all devices You can connect them over USB or you can use Universal Control wirelessly as long as the devices are within feet of each other	2021-12-13 20:27:44
Cisco	Cisco Blog	How Cisco IT is solving multi-cloud management: a single pane of glass	https://blogs.cisco.com/ciscoit/how-cisco-it-is-solving-multi-cloud-management-a-single-pane-of-glass	How Cisco IT is solving multi cloud management a single pane of glassWith continued growth in the consumption of cloud services enterprises now face the newest challenge ーmanaging multiple public and private cloud deployments Learn how Cisco IT approached the development of its solution for managing multi cloud controlling costs and giving internal customers greater visibility and control	2021-12-13 20:13:06
海外科学	NYT > Science	As Vaccines Trickle into Africa, Zambia’s Challenges Highlight Other Obstacles	https://www.nytimes.com/2021/12/11/health/covid-vaccine-africa.html	As Vaccines Trickle into Africa Zambia s Challenges Highlight Other ObstaclesVaccinating Africa is critical to protecting the continent and the world against dangerous variants but supply isn t the only problem countries face	2021-12-13 20:05:17
ニュース	BBC News - Home	Covid: First UK death recorded with Omicron variant	https://www.bbc.co.uk/news/uk-59639007?at_medium=RSS&at_campaign=KARANGA	booster	2021-12-13 20:25:09
ニュース	BBC News - Home	Harry Dunn crash: Anne Sacoolas case to go before UK court	https://www.bbc.co.uk/news/uk-england-northamptonshire-59643750?at_medium=RSS&at_campaign=KARANGA	harry	2021-12-13 20:51:02
ニュース	BBC News - Home	Premier League reports 42 positive Covid-19 results in past week	https://www.bbc.co.uk/sport/football/59643714?at_medium=RSS&at_campaign=KARANGA	Premier League reports positive Covid results in past weekForty two Premier League players and staff tested positive for Covid in the past week the most results ever recorded in the league over a seven day period	2021-12-13 20:47:19
ビジネス	ダイヤモンド・オンライン - 新着記事	住宅メーカー「領土拡大戦争」が激化！オープンハウスは大阪参戦、地方王者は首都圏侵攻 - 戸建てバブルの裏側	https://diamond.jp/articles/-/290016	虎視眈々	2021-12-14 05:25:00
ビジネス	ダイヤモンド・オンライン - 新着記事	薬剤師「年収と出世」のリアル！年収450万円で激務の病院、再就職できない製薬MR… - 薬剤師31万人 薬局6万店の大淘汰	https://diamond.jp/articles/-/290044	調剤薬局	2021-12-14 05:20:00
ビジネス	ダイヤモンド・オンライン - 新着記事	みずほ社外取がトップ解任を最後まで迷った理由、社長が営業店に残した「劇薬」の正体 - みずほ 退場宣告	https://diamond.jp/articles/-/289242		2021-12-14 05:15:00
ビジネス	ダイヤモンド・オンライン - 新着記事	薬学部「淘汰危険度」ランキング【55私立大】2位千葉科学大、1位は？ - 薬剤師31万人 薬局6万店の大淘汰	https://diamond.jp/articles/-/290043	の私立大薬学部の「淘汰危険度」ランキングを作成し、生き残る大学の条件を探った。	2021-12-14 05:10:00
ビジネス	ダイヤモンド・オンライン - 新着記事	JR東海は最終赤字に下方修正…コロナ禍2年目でも被害甚大な鉄道各社の事情 - ダイヤモンド 決算報	https://diamond.jp/articles/-/290518		2021-12-14 05:05:00
ビジネス	電通報 | 広告業界動向とマーケティングのコラム・ニュース	OOHの統合メディアプランニング	https://dentsu-ho.com/articles/8009	oohoutofhome	2021-12-14 06:00:00
ビジネス	電通報 | 広告業界動向とマーケティングのコラム・ニュース	DXの鍵を握るのは出産や育児でキャリアを離れた“潜在人材”だ！	https://dentsu-ho.com/articles/8008	表裏	2021-12-14 06:00:00
ビジネス	電通報 | 広告業界動向とマーケティングのコラム・ニュース	若い人の可能性を芽吹かせる「人事」とは？	https://dentsu-ho.com/articles/7929	八木洋介	2021-12-14 06:00:00
北海道	北海道新聞	住宅火災、男女２人死亡　夫婦か、けが人も、東京	https://www.hokkaido-np.co.jp/article/622619/	東京都中野区沼袋	2021-12-14 05:12:00
北海道	北海道新聞	＜社説＞衆院予算委論戦　目指す社会像見えたか	https://www.hokkaido-np.co.jp/article/622574/	予算委員会	2021-12-14 05:02:00
ビジネス	東洋経済オンライン	何でも｢レンチン調理｣への痛烈な違和感､5大欠点 作ってみて驚いた｢本当においしいですか？｣ | グルメ・レシピ | 東洋経済オンライン	https://toyokeizai.net/articles/-/475221?utm_source=rss&utm_medium=http&utm_campaign=link_back	東洋経済オンライン	2021-12-14 05:30:00
GCP	Cloud Blog	Find anything blazingly fast with Google's vector search technology	https://cloud.google.com/blog/topics/developers-practitioners/find-anything-blazingly-fast-googles-vector-search-technology/	Find anything blazingly fast with Google x s vector search technologyRecently Google Cloud partner Groovenauts Inc published a live demo of MatchIt Fast As the demo shows you can find images and text similar to a selected sample from a collection of millions in a matter of milliseconds Give it a try ーand either select a preset image or upload one of your own Once you make your choice you will get the top similar images from two million images on Wikimedia images in an instant as you can see in the video above No caching involved The demo also lets you perform the similarity search with news articles Just copy and paste some paragraphs from any news article and get similar articles from million articles on the GDELT project within a second Vector Search the technology behind Google Search YouTube Play and moreHow can it find matches that fast The trick is that the MatchIt Fast demo uses the vector similarity search or nearest neighbor search or simply vector search capabilities of the Vertex AI Matching Engine which shares the same backend as Google Image Search YouTube Google Play and more for billions of recommendations and information retrievals for Google users worldwide The technology is one of the most important components of Google s core services and not just for Google it is becoming a vital component of many popular web services that rely on content search and information retrieval accelerated by the power of deep neural networks So what s the difference between traditional keyword based search and vector similarity search For many years relational databases and full text search engines have been the foundation of information retrieval in modern IT systems For example you would add tags or category keywords such as movie music or actor to each piece of content image or text or each entity a product user IoT device or anything really You d then add those records to a database so you could perform searches with those tags or keywords In contrast vector search uses vectors where each vector is a list of numbers for representing and searching content The combination of the numbers defines similarity to specific topics For example if an image or any content includes of “movie of “music and of “actor related content then you could define a vector to represent it Note this is an overly simplified explanation of the concept the actual vectors have much more complex vector spaces You can find similar content by comparing the distances and similarities between vectors This is how Google services find valuable content for a wide variety of users worldwide in milliseconds With keyword search you can only specify a binary choice as an attribute of each piece of content it s either about a movie or not either music or not and so on Also you cannot express the actual meaning of the content to search If you specify a keyword films for example you would not see any content related to movies unless there was a synonyms dictionary that explicitly linked these two terms in the database or search engine  Vector search provides a much more refined way to find content with subtle nuances and meanings Vectors can represent a subset of content that contains much about actors some about movies and a little about music Vectors can represent the meaning of content where “films “movies and “cinema are all collected together Also vectors have the flexibility to represent categories previously unknown to or undefined by service providers For example emerging categories of content primarily attractive to kids such as ASMR or slime are really hard for adults or marketing professionals to predict beforehand and going back through vast databases to manually update content with these new labels would be all but impossible to do quickly But vectors can capture and represent never before seen categories instantly Vector search changes businessVector search is not only applicable to image and text content It can also be used for information retrieval for anything you have in your business when you can define a vector to represent each thing Here are a few examples Finding similar users If you define a vector to represent each user in your business by combining the user s activities past purchase history and other user attributes then you can find all users similar to a specified user  You can then see for example users who are purchasing similar products users that are likely bots or users who are potential premium customers and who should be targeted with digital marketing Finding similar products or items With a vector generated from product features such as description price sales location and so on you can find similar products to answer any number of questions for example What other products do we have that are similar to this one and may work for the same use case or What products sold in the last hours in this area based on time and proximity Finding defective IoT devices With a vector that captures the features of defective devices from their signals vector search enables you to instantly find potentially defective devices for proactive maintenance Finding ads Well defined vectors let you find the most relevant or appropriate ads for viewers in milliseconds at high throughput Finding security threats You can identify security threats by vectorizing the signatures of computer virus binaries or malicious attack behaviors against web services or network equipment   and many more Thousands of different applications of vector search in all industries will likely emerge in the next few years making the technology as important as relational databases OK vector search sounds cool But what are the major challenges to applying the technology to real business use cases Actually there are two Creating vectors that are meaningful for business use casesBuilding a fast and scalable vector search serviceEmbeddings meaningful vectors for business use casesThe first challenge is creating vectors for representing various entities that are meaningful and useful for business use cases This is where deep learning technology can really shine In the case of the MatchIt Fast demo the application simply uses a pre trained MobileNet v model for extracting vectors from images and the Universal Sentence Encoder USE for text By applying such models to raw data you can extract embeddings vectors that map each row of data in a space of their meanings MobileNet puts images that have similar patterns and textures closer to one another in the embedding space and USE puts texts that have similar topics closer For example a carefully designed and trained machine learning model could map movies into an embedding space like the following With the embedding space shown here users could find recommended movies based on the two dimensions is the movie for children or adults and is it a blockbuster or arthouse movie This is a very simple example of course but with an embedding space like this that fits your business requirements you can deliver a better user experience on recommendation and information retrieval services with insights extracted from the model  For more about creating embeddings the Machine Learning Crash Course on Recommendation Systems is a great way to get started We will also discuss how to extract better embeddings from business data later in this post Building a fast and scalable vector search serviceSuppose that you have successfully extracted useful vectors embeddings from your business data Now the only thing you have to do is search for similar vectors That sounds simple but in practice it is not Let s see how the vector search works when you implement it with BigQuery in a naive way It takes about seconds to find similar items fish images in this case from a pool of one million items That level of performance is not so impressive especially when compared to the MatchIt Fast demo BigQuery is one of the fastest data warehouse services in the industry so why does the vector search take so long This illustrates the second challenge building a fast and scalable vector search engine isn t an easy task The most widely used metrics for calculating the similarity between vectors are L distance Euclidean distance cosine similarity and inner product dot product But all require calculations proportional to the number of vectors multiplied by the number of dimensions if you implement them in a naive way For example if you compare a vector with elements to M vectors the number of calculations will be proportional to x M B This is the computation required to look through all the entities for a single search and the reason why the BigQuery demo above takes so long Instead of comparing vectors one by one you could use the approximate nearest neighbor ANN approach to improve search times Many ANN algorithms use vector quantization VQ in which you split the vector space into multiple groups define codewords to represent each group and search only for those codewords This VQ technique dramatically enhances query speeds and is the essential part of many ANN algorithms just like indexing is the essential part of relational databases and full text search engines As you may be able to conclude from the diagram above as the number of groups in the space increases the speed of the search decreases and the accuracy increases  Managing this trade off ーgetting higher accuracy at shorter latency ーhas been a key challenge with ANN algorithms  Last year Google Research announced ScaNN a new solution that provides state of the art results for this challenge With ScaNN they introduced a new VQ algorithm called anisotropic vector quantization Anisotropic vector quantization uses a new loss function to train a model for VQ for an optimal grouping to capture farther data points i e higher inner product in a single group With this idea the new algorithm gives you higher accuracy at lower latency as you can see in the benchmark result below the violet line  This is the magic ingredient in the user experience you feel when you are using Google Image Search YouTube Google Play and many other services that rely on recommendations and search In short Google s ANN technology enables users to find valuable information in milliseconds in the vast sea of web content How to use Vertex AI Matching EngineNow you can use the same search technology that powers Google services with your own business data Vertex AI Matching Engine is the product that shares the same ScaNN based backend with Google services for fast and scalable vector search and recently it became GA and ready for production use In addition to ScaNN Matching Engine gives you additional features as a commercial product including Scalability and availability The open source version of ScaNN is a good choice for evaluation purposes but as with most new and advanced technologies you can expect challenges when putting it into production on your own For example how do you operate it on multiple nodes with high scalability availability and maintainability Matching Engine uses Google s production backend for ScaNN which provides auto scaling and auto failover with a large worker pool It is capable of handling tens of thousands of requests per second and returns search results in less than ms for the th percentile with a recall rate of Fully managed You don t have to worry about building and maintaining the search service Just create or update an index with your vectors and you will have a production ready ANN service deployed No need to think about rebuilding and optimizing indexes or other maintenance tasks Filtering Matching Engine provides filtering functionality that enables you to filter search results based on tags you specify on each vector For example you can assign country and stocked tags to each fashion item vector and specify filters like US OR Canada AND stocked  or not Japan AND stocked on your searches Let s see how to use Matching Engine with code examples from the MatchIt Fast demo Generating embeddingsBefore starting the search you need to generate embeddings for each item like this one This is an embedding with dimensions for a single image generated with a MobileNet v model The MatchIt Fast demo generates embeddings for two million images with the following code After you generate the embeddings you store them in a Google Cloud Storage bucket  Configuring an indexThen define aJSON file for the index configuration You can find a detailed description for each field in the documentation but here are some important fields contentsDeltaUri the place where you have stored the embeddingsdimensions how many dimensions in the embeddingsapproximateNeighborsCount the default number of neighbors to find via approximate search distanceMeasureType how the similarity between embeddings should be measured either L L cosine or dot product this page explains which one to choose for different embeddings To create an index on the Matching Engine run the following gcloud command where the metadata file option takes the JSON file name defined above Run the searchNow the Matching Engine is ready to run The demo processes each search request in the following order First the web UI takes an image the one chosen or uploaded by the user and encodes it into an embedding using the TensorFlow js MobileNet v model running inside the browser Note this client side encoding is an interesting option for reducing network traffic when you can run the encoding at the client In many other cases you would encode contents to embeddings with a server side prediction service such as Vertex AI Prediction or just retrieve pre generated embeddings from a repository like Vertex AI Feature Store The App Engine frontend receives the embedding and submits a query to the Matching Engine Note that you can also use any other compute services in Google Cloud for submitting queries to Matching Engine such as Cloud Run Compute Engine or Kubernetes Engine or whatever is most suitable for your applications Matching Engine executes its search The connection between App Engine and Matching Engine is provided via a VPC private network for optimal latency Matching Engine returns the IDs of similar vectors in its index Step is implemented with the following code The request to the Matching Engine is sent via gRPC as you can see in the code above After it gets the request object it specifies the index id appends elements of the embedding specifies the number of neighbors similar embeddings to retrieve and calls the Match function to send the request The response is received within milliseconds Next steps Making changes for various use cases and better search qualityAs we noted earlier the major challenges in applying vector search on production use cases are Creating vectors that are meaningful for business use casesBuilding a fast and scalable vector search serviceFrom the example above you can see that Vertex AI Matching Engine solves the second challenge What about the first one Matching Engine is a vector search service it doesn t include the creating vectors part The MatchIt Fast demo uses a simple way of extracting embeddings from images and contents specifically it uses an existing pre trained model either MobileNet v or Universal Sentence Encoder While those are easy to get started with you may want to explore other options to generate embeddings for other use cases and better search quality based on your business and user experience requirements For example how do you generate embeddings for product recommendations  The Recommendation Systems section of the Machine Learning Crash Course is a great resource for learning how to use collaborative filtering and DNN models the two tower model to generate embeddings for recommendation Also TensorFlow Recommenders provides useful guides and tutorials for the topic especially on the two tower model and advanced topics For integration with Matching Engine you may also want to check out the Train embeddings by using the two tower built in algorithm page Another interesting solution is the Swivel model Swivel is a method for generating item embeddings from an item co occurrence matrix For structured data such as purchase orders the co occurrence matrix of items can be computed by counting the number of purchase orders that contain both product A and product B for all products you want to generate embeddings for To learn more take a look at this tutorial on how to use the model with Matching Engine If you are looking for more ways to achieve better search quality consider metric learning which enables you to train a model for discrimination between entities in the embedding space not only classification Popular pre trained models such as the MobileNet v can classify each object in an image but they are not explicitly trained to discriminate the objects from each other with a defined distance metric With metric learning you can expect better search quality by designing the embedding space optimized for various business use cases TensorFlow Similarity could be an option for integrating metric learning with Matching Engine Interested Today we re just beginning the migration from traditional search technology to new vector search Over the next to years many more best practices and tools will be developed in the industry and community These tools and best practices will help answer many questions like How do you design your own embedding space for a specific business use case How do you measure search quality How do you debug and troubleshoot the vector search How do you build a hybrid setup with existing search engines for meeting sophisticated requirements There are many new challenges and opportunities ahead for introducing the technology to production Now s the time to get started delivering better user experiences and seizing new business opportunities with Matching Engine powered by vector search AcknowledgementsWe would like to thank Anand Iyer Phillip Sun and Jeremy Wortz for their invaluable feedback to this post Related ArticleVertex Matching Engine Blazing fast and massively scalable nearest neighbor searchSome of the handiest tools in an ML engineer s toolbelt are vector embeddings a way of representing data in a dense vector space  An ear Read Article	2021-12-13 20:45:00