投稿時間:2022-09-17 03:29:42 RSSフィード2022-09-17 03:00 分まとめ(31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
AWS	AWS Marketplace	Unified authorization for AWS with Styra Declarative Authorization Service: EKS Edition	https://aws.amazon.com/blogs/awsmarketplace/unified-authorization-aws-styra-declarative-authorization-service-eks-edition/	Unified authorization for AWS with Styra Declarative Authorization Service EKS EditionImplement EKS Guardrails with Open Policy Agent and Styra DAS How do AWS users control which people and machines can perform which actions within their custom applications and within the stack of development tools needed to build and run those applications Take for instance Kubernetes users of AWS Elastic Kubernetes Service EKS need comprehensive controls …	2022-09-16 17:20:37
AWS	AWS The Internet of Things Blog	Managing Organizational Transformation for Successful OT/IT Convergence	https://aws.amazon.com/blogs/iot/managing-organizational-transformation-for-successful-ot-it-convergence/	Managing Organizational Transformation for Successful OT IT ConvergenceIntroduction Industrial organizations are facing a new challenge as they try to merge the traditional physical world Operational Technology or OT and the digital world Information Technology or IT In our experience companies who prioritize organizational change management when implementing digital solutions get better results from their investments This is even more true when building …	2022-09-16 17:30:35
AWS	AWS Security Blog	154 AWS services achieve HITRUST certification	https://aws.amazon.com/blogs/security/154-aws-services-achieve-hitrust-certification/	AWS services achieve HITRUST certificationThe AWS HITRUST Compliance Team is excited to announce that Amazon Web Services AWS services are certified for the Health Information Trust Alliance HITRUST Common Security Framework CSF v for the cycle These AWS services were audited by a third party assessor and certified under the HITRUST CSF The full list is now …	2022-09-16 17:39:25
AWS	AWS Security Blog	154 AWS services achieve HITRUST certification	https://aws.amazon.com/blogs/security/154-aws-services-achieve-hitrust-certification/	AWS services achieve HITRUST certificationThe AWS HITRUST Compliance Team is excited to announce that Amazon Web Services AWS services are certified for the Health Information Trust Alliance HITRUST Common Security Framework CSF v for the cycle These AWS services were audited by a third party assessor and certified under the HITRUST CSF The full list is now …	2022-09-16 17:39:25
AWS	lambdaタグが付けられた新着投稿 - Qiita	Lambda で CloudWatch Logs のログを自動で S3 にアーカイブする、をちょこっと便利にした。	https://qiita.com/masa03/items/f1ffad14785381791a8d	cloudwatchlog	2022-09-17 02:23:10
AWS	AWSタグが付けられた新着投稿 - Qiita	Lambda で CloudWatch Logs のログを自動で S3 にアーカイブする、をちょこっと便利にした。	https://qiita.com/masa03/items/f1ffad14785381791a8d	cloudwatchlog	2022-09-17 02:23:10
海外TECH	MakeUseOf	Should You Share Your Netflix Password With Others?	https://www.makeuseof.com/tag/share-netflix-password-others/	netflix	2022-09-16 17:30:14
海外TECH	MakeUseOf	How to Open WPS (Works) Documents in Windows 10 & 11	https://www.makeuseof.com/windows-open-wps-files/	How to Open WPS Works Documents in Windows amp Microsoft Works may be no more but people still have WPS files left over from when they used it Here s how to open them in the modern day	2022-09-16 17:15:14
海外TECH	DEV Community	Introducing the Semantic Graph	https://dev.to/neuml/introducing-the-semantic-graph-1j09	Introducing the Semantic GraphThis article is part of a tutorial series on txtai an AI powered semantic search platform txtai executes machine learning workflows to transform data and build AI powered semantic search applications One of the main use cases of txtai is semantic search over a corpus of data Semantic search provides an understanding of natural language and identifies results that have the same meaning not necessarily the same keywords Within an Embeddings instance sits a wealth of implied knowledge and relationships between rows Many approximate nearest neighbor ANN indexes are even backed by graphs What if we are able to tap into this knowledge Semantic graphs also known as knowledge graphs or semantic networks build a graph network with semantic relationships connecting the nodes In txtai they can take advantage of the relationships inherently learned within an embeddings index This opens exciting possibilities for exploring relationships such as topics and interconnections in a dataset This article introduces the semantic graph Install dependenciesInstall txtai and all dependencies We ll install the graph extra for graph functionality pipeline extra for object detection and similarity extra to load models with the sentence transformers library Install txtaipip install txtai graph pipeline similarity datasets ipyplot Graph basicsFirst we ll build a basic graph and show how it can be used to explore relationships The code below builds a graph of animals and relationships between them We ll add nodes and relationships along with running a couple analysis functions import networkx as nxfrom txtai graph import GraphFactory Create graphgraph GraphFactory create backend networkx graph initialize Add nodesnodes dog fox wolf zebra horse labels uid text for uid text in nodes for uid text in nodes graph addnode uid text text Add relationshipsedges for source target weight in edges graph addedge source target weight weight Print centrality and path between and print Centrality labels k v for k v in graph centrality items print Path dog gt horse gt join labels uid for uid in graph showpath Visualize graphnx draw graph backend nx shell layout graph backend labels labels with labels True node size node color af edge color cfcfcf font color fff Centrality wolf dog fox zebra horse Path dog gt horse dog gt wolf gt zebra gt horseThe visualization shows the layout of the graph A centrality and path function were also run Centrality shows the most central or related nodes In this case the wolf node has the highest score We also ran a path function to show how the graph is traversed from dog to horse Build a Semantic GraphWhile txtai graphs can be standalone with nodes and relationships manually added the real power comes in indexing an embeddings instance The following section builds an embeddings index over the ag news dataset ag news contains news headlines from the mid s This configuration sets the familiar vector model and content settings Column expressions is a feature starting with txtai Column expressions alias expressions allowing SQL statements to use those references as a shorthand for the expression Next comes the graph The configuration sets the maximum number of connections to add per node along with a minimum similarity score Topic modeling parameters are also added which we ll cover later from datasets import load datasetfrom txtai embeddings import Embeddings Create embeddings instance with a semantic graphembeddings Embeddings path sentence transformers all MiniLM L v content True functions name graph function graph attribute expressions name category expression graph indexid category name topic expression graph indexid topic name topicrank expression graph indexid topicrank graph limit minscore topics categories Society amp Culture Science amp Mathematics Health Education amp Reference Computers amp Internet Sports Business amp Finance Entertainment amp Music Family amp Relationships Politics amp Government Load datasetdataset load dataset ag news split train rows dataset text Index datasetembeddings index x text None for x text in enumerate rows The embeddings index is now created Let s explore Topic modelingTopic modeling is an unsupervised method to identify abstract topics within a dataset The most common way to do topic modeling is to use clustering algorithms to group nodes with the closest proximity A number of excellent topic modeling libraries exist in Python today BERTopic and TopVec are two of the most popular Both use sentence transformers to encode data into vectors UMAP for dimensionality reduction and HDBSCAN to cluster nodes Given that an embeddings index has already encoded and indexed data we ll take a different approach txtai builds a graph running a query for each node against the index In addition to topic modeling this also opens up much more functionality which will be covered later Topic modeling in txtai is done using community detection algorithms Similar nodes are group together There are settings to control how much granularity is used to group nodes In other words topics can be very specific or broad depending on these settings Topics are labeled by building a BM index over each topic and finding the most common terms associated with the topic Let s take a closer look at the topics created with this embeddings index Store reference to graphgraph embeddings graphlen embeddings graph topics list graph topics keys kerry john bush president sox red boston series oil opec prices said dollar reuters against euro darfur sudan region said The section above shows the number of topics in the index and top topics Keep in mind that ag news is from the mid s and that is evident with the top topics Given that we added functions to run SQL functions to get the topic for each row we can use that to explore topics Each topic is associated with a list of associated matching ids Those ids are ranked based on the importance to the topic in a field named topicrank The section below prints the best matching text for the topic sox red boston series print embeddings search select text from txtai where topic sox red boston series and topicrank text Red Sox heading to the World Series The Boston Red Sox have won the American League Championship Series and are heading to the World Series for the first time since In addition to topics higher level categories can be associated with topics This enables having granular topics and encompassing categories for topics For example the topic of sox red boston series has a category of Sports See below for x topic in enumerate list graph topics keys print graph categories x topic Politics amp Government kerry john bush presidentSports sox red boston seriesBusiness amp Finance oil opec prices saidBusiness amp Finance dollar reuters against euroPolitics amp Government darfur sudan region saidTopics and categories can also be used to filter results See the difference when just querying for results similar to book and similar to book with a topic of Sports print embeddings search select text from txtai where similar book text A Guidebook for Every Taste LANNING a trip involves many difficult decisions but near the top of my list is standing in a bookstore trying to choose from a daunting lineup of guidebooks a purchase that brands the owner print embeddings search select text from txtai where category Sports and similar book text Same story for Wildcats After a game about as artful as a dime store novel Virginia coach Pete Gillen turned to literature to express the trying time his No Graph analysisIndexing an embeddings instance into a graph adds the ability to do network analysis For example the centrality of the graph can be analyzed to find the most common nodes Alternatively pagerank could also be run to rank the importance of nodes within the dataset The section below runs graph centrality and shows the associated topic for the most central nodes Not surprisingly many of the topics are top topics centrality graph centrality topics list graph topics keys for uid in list centrality keys topic graph attribute uid topic print f topic topics index topic peoplesoft oracle takeover bid darfur sudan region said windows microsoft xp service fallujah us city iraqi eclipse lunar moon total Walk the graphGiven that graphs are nodes and relationships we can traverse the nodes using those relationships The graph can be used to show how any two nodes are connected from IPython display import HTMLdef highlight index result output f index spans token score fffd if score gt else None for token score in result tokens if result score gt and not color for color in spans if color mscore max score for score in spans spans token score fffd if score mscore else color for token score color in spans for token color in spans output f lt span style background color color gt token lt span gt if color else f token return outputdef showpath source target path graph showpath source target path graph attribute p text for p in path sections for x p in enumerate path if x Print start node sections append f x p if x lt len path Explain and highlight next path element results embeddings explain p path x limit sections append highlight x results return HTML lt br gt lt br gt join sections showpath This shows how text about a famous squirrel and the Red Sox winning the world series are connected Notice how the first match pivots to a node about a squirrel running on the field during a baseball game From there it s a relatively logical path to the end node This is reminiscent of the game six degrees of Kevin Bacon Try running showpath with calls to random randint len rows it s oddly addicting This is a fun way to explore the interconnectivity of a dataset Group images into topicsTopic modeling isn t limited to text It supports any data that can be vectorized into an embeddings index Next we ll create an embeddings index using the imagenette dataset which is a small dataset for image object detectiondataset load dataset frgfm imagenette px split train rows dataset image Index with content and objectsembeddings Embeddings method sentence transformers path sentence transformers clip ViT B content True objects image functions name graph function graph attribute expressions name topic expression graph indexid topic name topicrank expression graph indexid topicrank graph limit minscore topics resolution embeddings index x image None for x image in enumerate rows graph embeddings graphlist graph topics keys topic topic topic topic topic Topic labeling for imagesThe index is now ready Notice how the topic names are generic Given there is no text associated a different approach is needed We ll use an object detection pipeline to label to best matching image per topic import ipyplotfrom PIL import Imagefrom txtai pipeline import Objectsdef labels objects Objects classification True threshold images embeddings search select topic object from txtai where topicrank order by topic results objects result object for result in images flatten True return images topic results x split for x images in enumerate images def scale image factor width height image size return image resize int width factor int width factor images labels labels for topic in list graph topics keys for result in embeddings search f select topic object from txtai where topic topic and topicrank len graph topics images topic scale result object ipyplot plot images list images values labels topic for topic in images img width As we can see each topic is now labeled Imagenette is a labeled dataset let s evaluate the accuracy of our topic modeling def accuracy correct total labels dataset label for topic in graph topics label labels int graph topics topic correct sum if labels int x label else for x in graph topics topic total len graph topics topic print Accuracy correct total accuracy Accuracy Not bad accuracy using a totally unsupervised method not even intended for image classification Walk the image graphAs we did before let s walk the graph We ll start with two images a person parachuting from the sky and someone holding a french horn images for uid in graph showpath images append scale embeddings search f select object from txtai where indexid uid limit object ipyplot plot images images img width Very interesting The first match is a person parachuting onto a football field followed by a matching band on a field finally leading to a person holding a french horn Wrapping upThis article covered quite a lot We introduced graphs showed how they can be used to model semantic relationships and topics This change makes it easier to run exploratory data analysis on a dataset with txtai and quickly gain insights This is just the beginning of what is possible and there are a wide range of exciting new possibilities for txtai stay tuned	2022-09-16 17:24:55
海外TECH	DEV Community	Javascript vs Ruby: Syntax Edition	https://dev.to/adenaddis/javascript-vs-ruby-syntax-edition-3mhp	Javascript vs Ruby Syntax EditionTransitioning from consistently working with javascript to moving straight into ruby might be difficult due to adjusting our eyes to look at a new way of writing code and with that changing our habits Lets start simple Javascript Syntaxfunction jsFunction param console log Look how Im written return param Lets jot down a couple things we see with js syntax Use starter keyword function to show the code is a functionjsFunction is the variable name we use this term to refer to the code wherever we please NEED parentheses right after the variable name to place any parameters we might have in them param is the variable name for the functions parameter that will be used when the function is later passing an argument and being invoked Next curly brackets is a MUST to indicate the body of the functionconsole log is what will be displayed in the terminalreturn is different in which it is not specifically for the terminal We use this so that our function can have a return value when it is called on For this example we have param plus How to check function const jsFunctionReturnValue jsFunction gt Look how Im written console log myFunctionReturnValue gt Above we just called the jsFunction and passed an argument of and assigned that to a variable called jsFunctionReturnValue This in turn displays my earlier console log in the terminal When I console log the new variable I made I get the number due to the Ruby Syntaxdef rb method param puts Look how Im written param endLets see what makes ruby different Use starter keyword def to show the code is a methodWe use snake case for the variable name of the method This just means the words are separated by an underscore While JS uses camel case Parameters still come after the method name parentheses are optionalSince we don t need curly braces here like in js we use the keyword end after the code to identify where the code finishes The return keyword is not needed in Ruby but you can use it It is just known to ruby that the last line will the the return value Now to check Ruby code you need to run IRB and put your previous code press enter then check with the following rb method return value rb method Look how Im written gt rb method return value gt Here we can see the argument of being passed into our method and being assigned to a new variable which then we can call on and we get because of Fun fact	2022-09-16 17:14:16
Apple	AppleInsider - Frontpage News	Deals: get an iPhone 14, iPhone 14 Pro for free with wireless carrier promos	https://appleinsider.com/articles/22/09/09/deals-get-an-iphone-14-iphone-14-pro-for-free-with-wireless-carrier-promos?utm_medium=rss	Deals get an iPhone iPhone Pro for free with wireless carrier promosSnag yourself an iPhone at a discount or save money on an unlimited data plan by checking out U S carriers latest promotional offers iPhone deals abound through wireless carriers including AT amp T and Verizon Verizon Read more	2022-09-16 17:44:34
海外TECH	Engadget	'Hollow Knight: Silksong' will come to the PS4 and PS5, eventually	https://www.engadget.com/hollow-knight-silksong-ps4-ps5-team-cherry-172014333.html?src=rss	x Hollow Knight Silksong x will come to the PS and PS eventuallyHollow Knight nbsp fans who tend to only play games on PlayStation can breathe a sigh of relief The long awaited sequel Hollow Knight Silksong will be coming to PlayStation and PlayStation The game had already been confirmed for Nintendo Switch Xbox PC Mac and Linux Silksong is also coming to Xbox Game Pass PC Game Pass and Xbox Cloud Gaming on its release day Only trouble is we still don t know exactly when that will be Xbox indicated during its June showcase that Silksong would be out in the following year so the release date will likely be sometime in the next nine months Here s hoping developer and publisher Team Cherry reveals that date very soon Sharpen your needles confirming Hollow Knight Silksong is coming to PS and PS pic twitter com poIclQDfvrーPlayStation PlayStation September	2022-09-16 17:20:14
ニュース	BBC News - Home	Queue for Queen's lying-in-state reopens after seven hours	https://www.bbc.co.uk/news/uk-62933685?at_medium=RSS&at_campaign=KARANGA	resumes	2022-09-16 17:21:05
ニュース	BBC News - Home	King Charles III in Cardiff for first Wales visit as monarch	https://www.bbc.co.uk/news/uk-wales-62915136?at_medium=RSS&at_campaign=KARANGA	death	2022-09-16 17:29:21
ニュース	BBC News - Home	Watchdog to examine race as a factor in police shooting	https://www.bbc.co.uk/news/uk-england-london-62921649?at_medium=RSS&at_campaign=KARANGA	chris	2022-09-16 17:26:47
ニュース	BBC News - Home	Transport bosses warn mourners of London queues	https://www.bbc.co.uk/news/business-62930532?at_medium=RSS&at_campaign=KARANGA	elizabeth	2022-09-16 17:42:27
ニュース	BBC News - Home	David Beckham queues for 12 hours to see Queen lying in state	https://www.bbc.co.uk/sport/football/62932806?at_medium=RSS&at_campaign=KARANGA	london	2022-09-16 17:26:01
ニュース	BBC News - Home	Queen’s funeral: Community volunteers get ‘surprise’ invitations	https://www.bbc.co.uk/news/uk-62932708?at_medium=RSS&at_campaign=KARANGA	choir	2022-09-16 17:40:19
ニュース	BBC News - Home	Which businesses will close and stay open on day of Queen's funeral?	https://www.bbc.co.uk/news/business-62879563?at_medium=RSS&at_campaign=KARANGA	holiday	2022-09-16 17:03:51
ビジネス	ダイヤモンド・オンライン - 新着記事	なぜ「ミニマル化」は「サイフ」から始めたほうがいいと断言できるのか？ - 超ミニマル主義	https://diamond.jp/articles/-/309817	なぜ「ミニマル化」は「サイフ」から始めたほうがいいと断言できるのか超ミニマル主義「手放し、効率化し、超集中」するための全技法とはベストセラー『自由であり続けるために代で捨てるべきのこと』以来、四角大輔が年ぶりに書き下ろしたビジネス書『超ミニマル主義』の中から、「サイフ」「カバン」「書類」「名刺」「ウェア」「シューズ」「仕事机」「デバイス」「部屋」といった物質、「情報」「データ」「スケジュール」「タスク」「労働時間」「ストレス」「人付き合い」といった非物質を、極限まで「最小・最軽量化」する方法を紹介していく。	2022-09-17 02:55:00
ビジネス	ダイヤモンド・オンライン - 新着記事	天才・小室直樹が見ていた「日本社会の構造」とは - 【新装版】危機の構造　日本社会崩壊のモデル	https://diamond.jp/articles/-/309926	『【新装版】危機の構造日本社会崩壊のモデル』では、社会学者・橋爪大三郎氏による解説に加え、年に発刊された【増補版】に掲載された「私の新戦争論」も収録されている。	2022-09-17 02:50:00
ビジネス	ダイヤモンド・オンライン - 新着記事	「経歴詐称」で誰が実害を被るのか？ - ぼくらは嘘でつながっている。	https://diamond.jp/articles/-/309885	人間関係	2022-09-17 02:45:00
ビジネス	ダイヤモンド・オンライン - 新着記事	コミュ力の高い人は「悪口を言われたとき」にどう返す？ - おもろい話し方	https://diamond.jp/articles/-/309763	高い	2022-09-17 02:40:00
ビジネス	ダイヤモンド・オンライン - 新着記事	【人間関係の闇から抜け出す方法】高圧的な人を無意識に引き寄せる、危険すぎる考え方 - とても傷つきやすい人が無神経な人に悩まされずに生きる方法	https://diamond.jp/articles/-/309589		2022-09-17 02:35:00
ビジネス	ダイヤモンド・オンライン - 新着記事	【精神科医が教える】同じ苦労話を何度も話してしまう人の切実な理由 - こころの葛藤はすべて私の味方だ。	https://diamond.jp/articles/-/309910	精神科医	2022-09-17 02:30:00
ビジネス	ダイヤモンド・オンライン - 新着記事	【“FIRE達成”の大人気FPが解説！】 会社員が“退職する前”に調査しておきたい「お金にかかわる社内の制度」 - 年収300万円からのFIRE入門	https://diamond.jp/articles/-/309591	達成	2022-09-17 02:25:00
ビジネス	ダイヤモンド・オンライン - 新着記事	【中学受験に役立つ】 プログラミング教育の必修化、その本当の狙いとは？ - 中学受験を目指す保護者からよく質問される「子育てQ&A」	https://diamond.jp/articles/-/309722	【中学受験に役立つ】プログラミング教育の必修化、その本当の狙いとは中学受験を目指す保護者からよく質問される「子育てQampampA」開成・麻布・筑波大駒場・渋谷幕張…。	2022-09-17 02:20:00
ビジネス	ダイヤモンド・オンライン - 新着記事	【これがNo.1】トップセールス×メンタリズム＝最強のコミュニケーションスキル - メンタリズム日本一が教える「8秒」で人の心をつかむ技術	https://diamond.jp/articles/-/309925	【これがNo】トップセールス×メンタリズム最強のコミュニケーションスキルメンタリズム日本一が教える「秒」で人の心をつかむ技術ただ話すだけなのに「頑張る」「疲れる」「気を使う」……。	2022-09-17 02:15:00
ビジネス	ダイヤモンド・オンライン - 新着記事	「10万人のフォロワーがいるけど、いいねがつかない人」の特徴 - ブログで５億円稼いだ方法	https://diamond.jp/articles/-/309957	特徴	2022-09-17 02:10:00
ビジネス	ダイヤモンド・オンライン - 新着記事	【意識が及ぼす影響】苦しみと幸福の間で揺れる人間という存在 - ダマシオ教授の教養としての「意識」	https://diamond.jp/articles/-/309924	認識	2022-09-17 02:05:00
北海道	北海道新聞	ウポポイでアイヌ民族特別展　１７日から　重要文化財や道内初公開も	https://www.hokkaido-np.co.jp/article/732458/	胆振管内	2022-09-17 02:11:33