投稿時間:2022-11-18 20:20:21 RSSフィード2022-11-18 20:00 分まとめ(22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 Japan Blog AWS Toolkit for .NET Refactoring で.NET のモダナイゼーションを加速させる https://aws.amazon.com/jp/blogs/news/aws-toolkit-for-net-refactoring-launch/ ratenetmodernizatiovisun 2022-11-18 10:56:22
python Pythonタグが付けられた新着投稿 - Qiita 【Python】挿入ソートを盛大に勘違いしてしまった件 https://qiita.com/syoshika_/items/71b7c94dba2f174ea0e3 並べ替え 2022-11-18 19:57:22
技術ブログ Mercari Engineering Blog 得意領域を超えるためのフロントエンド勉強会を実施した際の工夫 https://engineering.mercari.com/blog/entry/20221117-front-end-study-session/ hellip 2022-11-18 11:00:57
海外TECH DEV Community In One Minute : Maven https://dev.to/rakeshkr2/in-one-minute-maven-45o In One Minute MavenApache Maven is a build automation tool used primarily for Java projects Based on the concept of a Project Object Model POM Maven can manage a project s build reporting and documentation from a central piece of information Maven s objectivesMaven s primary goal is to allow a developer to comprehend the complete state of a development effort in the shortest period of time Maven adheres to the principle of convention over configuration that helps building Maven projects with very little configuration Systems libraries and frameworks should assume reasonable defaults and systems should just work without requiring unnecessary configuration In order to attain this goal there are several areas of concern that Maven attempts to deal with Making the build process easyProviding a uniform build systemProviding quality project informationProviding guidelines for best practices developmentAllowing transparent migration to new featuresThe configuration of a Maven project is done by creating a file called pom xml located at the root of the project This file is referred to as the POM file Each POM file inherits from a global Super POM that defined common properties for all Maven projects More informationMaven Releases History with documentationApache Maven Wikipedia Article Free Maven BooksBetter Builds with MavenSonatype eBooks 2022-11-18 10:30:00
海外TECH DEV Community Tailwind CSS tutorial #15: Font Family https://dev.to/shubhicodes/tailwind-css-tutorial-15-font-family-fee Tailwind CSS tutorial Font FamilyIn the article we will go into detail on how to use font family Font FamilyFormatflex sans serif mono AlignmentTailwind ClassCSS PropertySansfont sansfont family ui sans serif system ui apple system BlinkMacSystemFont Segoe UI Roboto Helvetica Neue Arial Noto Sans sans serif Apple Color Emoji Segoe UI Emoji Segoe UI Symbol Noto Color Emoji Serifflex row reversefont family ui serif Georgia Cambria Times New Roman Times serif Monoflex colfont family ui monospace SFMono Regular Menlo Monaco Consolas Liberation Mono Courier New monospace let s see each of this in action Font SansThis class is used to apply a websafe sans serif font family like ui sans serif system ui apple system BlinkMacSystemFont “Segoe UI“ Roboto “Helvetica Neue“ Arial “Noto Sans“ sans serif “Apple Color Emoji“ “Segoe UI Emoji“ “Segoe UI Symbol“ “Noto Color Emoji“ etc lt li class flex items center justify between px py gt lt div gt lt div class font sans font light text xl align middle mb gt Font Sans lt div gt lt p class font sans text md leading tight gt Rex is the Mayor of Binaryville and a well loved personality in town He rose to robotdom from a microprocessor plant on the south side of town where many famous and influential robots before him were conceived lt p gt lt div class text xs font mono font light leading tight text gray mt gt lt span class font bold gt font family lt span gt ui sans serif system ui apple system BlinkMacSystemFont Segoe UI Roboto Helvetica Neue Arial Noto Sans sans serif Apple Color Emoji Segoe UI Emoji Segoe UI Symbol Noto Color Emoji lt div gt lt div gt lt div class text xs font semibold font mono whitespace nowrap px py ml rounded text white bg pink rounded gt font sans lt div gt lt li gt Output Font SerifThis class is used to apply a websafe serif font family like ui serif Georgia Cambria “Times New Roman“ Times serif etc lt li class flex items center justify between px py gt lt div gt lt div class font serif font light text xl align middle mb gt Font Serif lt div gt lt p class font serif text md leading tight gt Dolores robot is a computational genius who performs calculations for various Binaryville businesses and organizations In her spare time Dolores has been noodling on the creation of the perfect web browser that will make all other browsers obsolete We re rooting for you Dolores lt p gt lt div class text xs font mono font light leading tight text gray mt gt lt span class font bold gt font family lt span gt ui serif Georgia Cambria Times New Roman Times serif lt div gt lt div gt lt div class text xs font semibold font mono whitespace nowrap px py ml rounded text white bg pink rounded gt font serif lt div gt lt li gt Output Font MonoThis class is used to apply a websafe monospaced font family like ui monospace SFMono Regular Menlo Monaco Consolas “Liberation Mono“ “Courier New“ monospace etc lt li class flex items center justify between px py gt lt div gt lt div class font mono font light text xl align middle mb gt Font Mono lt div gt lt p class font mono text md leading tight gt Bubble Gum robot or “Bubbles for short is from a family of track footed robots that ¬originated from an experiment melding candy vending machines with robotics in the early s Bubbles is a favorite of Binaryville not just because she generously dispenses candy but also because she has one of the more bubbly personalities of the villagers lt p gt lt div class text xs font mono font light leading tight text gray mt gt lt span class font bold gt font family lt span gt ui monospace SFMono Regular Menlo Monaco Consolas Liberation Mono Courier New monospace lt div gt lt div gt lt div class text xs font semibold font mono whitespace nowrap px py ml rounded text white bg pink rounded gt font mono lt div gt lt li gt OutputFull code The overall code will be attached to repo link Overall OutputResources tailwind cssThank you for reading To learn more check out my blogs on Flex Direction Hackathons and Flex Wrap If you liked this article consider following me on Dev to for my latest publications You can reach me on Twitter Keep learning Keep coding 2022-11-18 10:17:19
海外TECH DEV Community Using Rossum.ai to superpower OCR with machine learning https://dev.to/finnauto/using-rossumai-to-superpower-ocr-with-machine-learning-b70 Using Rossum ai to superpower OCR with machine learning IntroAt FINN we strongly believe in the principle of do it do it automate it If you find yourself doing something manually twice you should find a way to automate the process This holds just as true when it comes to the precious details of invoices and financial document processing Within the FINN Finance team we ve recently started using a platform called Rossum that helps us improve the automatic optical character recognition OCR of documents But to understand why Rossum is so helpful let s first dive into the problems you ll frequently run into with traditional OCR tools Problems with traditional OCRConsider the sheer variety of invoices we receive on a daily basis In these invoice PDF snippets the fields indicating the total amount due are very different from one another Sometimes the currency is specified in euros in other cases it s in dollars Some invoices speak of Total amount or Balance due while still others use Brutto The formatting is all over the place which wreaks havoc on regex based OCR For humans such variation is relatively easy to read When a human looks at this they should not have much trouble distinguishing what is what A human would usually skim the documents and see that invoice number is in one place the date in another and so on In skimming a document like that we know to expect localized data i e the total should be somewhere under the itemized table Humans can also make quick decisions For example when you know dates only come in a certain format then you can easily pick up cases where there s something wrong with an invoice such as where the due date of the invoice appears to be before its issue date But such an approach is less straightforward for a machine And this is exactly where traditional OCR systems can face challenges When there is lots of variation in the input formatting traditional OCR might give you accurate data sure But it s not obvious that in all cases it will find exactly what you re looking for When you ve got tens of thousands of invoices to process that s not the sort of issue you want to run into So how did we at FINN solve this problem Superpowered OCR with machine learningEnter Rossum Rossum is a platform for as they call it intelligent document processing that automates the processing of variable documents And variable documents here means that you are dealing with documents that have very similar data such as data for the total amount due but in totally different formats different labels currencies date formatting and so on Rossum uses machine learning to improve and speed up the processing of such documents And it does so by trying to approach these invoice documents a bit more the way a human might So how does that work At FINN our use case is processing incoming invoices To do this we first set up a schema for all the data points that we want to collect from the invoice and train the relevant model After that initial training by and large it can run entirely automated though we might improve the training once in a while Next we set up our email automation so that we process invoices as soon as we get them We can define a confidence score with options ranging from to which indicates the degree of confidence of the AI engine that it has correctly identified the location and text of a field For instance if you set the required confidence score at then when the system is able to meet that threshold the whole OCR processing runs automatically and no human will have to look at it Only when the system is not able to meet the selected accuracy levels it will send you a notification so that a human can still review it Two parts make the process especially fun First because Rossum uses machine learning the system progressively gets better If you have a field that initially needed regular reviews then once you get enough manual reviews the system will learn to detect that field As a result the next time we see a similar kind of case it may very well hit the required accuracy levels and automate the entire process That s one huge advantage of using Rossum The other fun part is data validation It is super easy to set things up to validate a range of data For example to check that the supplier exists that the supplier s International Bank Account Number IBAN is in such and such format or even to catch plain errors We once had a supplier send us an invoice dated the st of June which is an impossible date Due to proper data validation it s very easy to catch that Traditional OCR does not work like this at all But such an approach is very intuitive to humans and Rossum is trying to approximate that Customization and training your own engineDifferent organizations will have different needs in terms of document processing At FINN we work with car subscriptions so it s often crucial for us to find things such as the car s license plate or the Vehicle Identification Number VIN on the invoices Rossum has a very good generic engine that we can use straightaway This already helps us find many of the relevant issues But in addition you can also purchase an engine that you can train for customized options that are specifically relevant to your organization So at FINN we decided to add things such as the VIN as specific options to the training process to tailor it to our specific needs The Rossum website also offers a bunch of different settings for customization It also allows us to distinguish between different types of documents to process such as invoices or speeding tickets Or you also have settings to specify what kind of language you re mostly working with Language processing is interesting Many conventions in financial documents differ between languages or from region to regionーthink of decimal separators or date formats You would think that that should result in quite some difficulties with OCR ing the invoices Well not really So far at FINN we haven t had to train separate models for say German versus US English invoices Rossum supports a ton of customization options it already covers most languages that use a variation of the Latin script and support for Japanese and Chinese language documents is currently in beta At FINN we are regularly processing invoices with a range of different formats In terms of really taking things further Rossum even allows you to add extensions with your own logic For example we may want to match the supplier name found on an invoice to the list of suppliers already known to FINN But in case of a new supplier you re obviously not going to find a match To solve this we built an app with our drag and drop internal tools building system Retool that we hooked into the Rossum web interface So with a single click you can go straight from Rossum to our dedicated Retool app and insert the new supplier Going yet further we already have a list of cost centers that we use on our finance management tool Once the system recognizes the supplier name we can also match this automatically to the cost center Once matched we can export and save all the information in our own database and upload it directly to our SaaS Finance Tool Ideally then once it s all trained the entire process is mostly automated ーit s all super smooth for the user AlternativesWe did evaluate a few other OCR tools like Nanonets and Super AI Rossum caters well to both non devs and devs The UI is very easy to use and they have great API Docs The options for custom extensions using Lambdas or Webhooks open up Rossum to great extent which makes up for some missing features We have built our own extensions for data validation detecting multiple invoices in a single pdf removal of irrelevant pages etc So that s how we at FINN are using Rossum to automate and speed up our financial document processing Have you had any challenges with financial document processing Let me know in the comments how you solved them Written by Chris Meyns Technical Writer and Gazala Muhamed Software Engineer Resources introduction 2022-11-18 10:07:27
海外TECH Engadget NASA picks SpaceX’s Starship for its second crewed Artemis lunar landing https://www.engadget.com/nasa-picks-spacex-artemis-4-102537076.html?src=rss NASA picks SpaceX s Starship for its second crewed Artemis lunar landingNASA has expanded SpaceX s role in the Artemis program and has selected the company s Starship lunar lander to ferry the second batch of astronauts to the Moon If you ll recall the agency picked SpaceX s human landing system for Artemis which will take humanity back to the Moon decades after the last Apollo mission nbsp Earlier this year NASA announced that it was accepting new lunar lander proposals for use beyond Artemis to ensure redundancy in services While SpaceX wasn t allowed to participate the agency did say that it was planning to exercise an option under their existing contract and was asking the company to modify its landing system to meet a new requirement That is for its lander to have the capability to take human spacefarers from the Gateway station which has yet to be installed in the lunar orbit to the Moon itself NASA can then use this upgraded lander for future missions as humanity attempts to establish a long term presence on the Moon nbsp NASA says in its announcement The aim of this new work under Option B is to develop and demonstrate a Starship lunar lander that meets NASA s sustaining requirements for missions beyond Artemis III including docking with Gateway accommodating four crew members and delivering more mass to the surface SpaceX s original contract with NASA was worth billion but this modification will add billion to the total The company s Starship lunar lander is expected to take astronauts to the Moon for the first time in However NASA Inspector General Paul Martin admitted in August that the agency is expecting Artemis s launch date to slip to Based on his statement NASA is anticipating delays related to the development of the human landing system and its next generation spacesuits NASA s Artemis mission which will take four more astronauts to the Moon will take flight in at the earliest nbsp For now we re still waiting for updates on the Orion vehicle after the successful Artemis launch on November th You can track the spacecraft s location by following its official Twitter account which frequently posts about its latest distance from Earth and from the Moon nbsp 2022-11-18 10:25:37
海外科学 NYT > Science Will Covid Boosters Prevent Another Wave? Scientists Aren’t So Sure. https://www.nytimes.com/2022/11/18/health/covid-boosters-surge.html Will Covid Boosters Prevent Another Wave Scientists Aren t So Sure The shots may help older pregnant and immunocompromised Americans dodge serious illness or death But the doses are not likely to prevent infections in any group recent studies suggest 2022-11-18 10:00:42
海外科学 BBC News - Science & Environment 'Landmark vote' gives boost to threatened sharks https://www.bbc.co.uk/news/science-environment-63670072?at_medium=RSS&at_campaign=KARANGA conservation 2022-11-18 10:28:58
医療系 医療介護 CBnews いつも相談する医師「いる」24%、健保連調べ-リフィル処方箋「通院中で利用希望」全体の3割 https://www.cbnews.jp/news/entry/20221118183341 健康保険組合連合会 2022-11-18 19:10:00
海外ニュース Japan Times latest articles Kyoto prosecutors indict gang member over murder of gyōza chain president https://www.japantimes.co.jp/news/2022/11/18/national/crime-legal/suspect-indicted-ohsho-president-murder/ chain 2022-11-18 19:02:04
海外ニュース Japan Times latest articles Co-leader Hoshoryu puts skills on display in impressive victory over Daieisho https://www.japantimes.co.jp/sports/2022/11/18/sumo/basho-reports/co-leader-hoshoryu-puts-skills-display-impressive-victory-daieisho/ Co leader Hoshoryu puts skills on display in impressive victory over DaieishoThe year old nephew of Mongolian great Asashoryu showed off his impressive strength and technique while improving to 2022-11-18 19:17:46
ニュース BBC News - Home Autumn Statement: Jeremy Hunt defends tax rises for middle earners https://www.bbc.co.uk/news/uk-politics-63674045?at_medium=RSS&at_campaign=KARANGA autumn 2022-11-18 10:53:10
ニュース BBC News - Home Mohammed bin Salman: Saudi leader given US immunity over Khashoggi killing https://www.bbc.co.uk/news/world-us-canada-63671947?at_medium=RSS&at_campaign=KARANGA minister 2022-11-18 10:13:04
ニュース BBC News - Home Strabane security alert: Police targeted in Mount Carmel Heights bomb attack https://www.bbc.co.uk/news/uk-northern-ireland-63674957?at_medium=RSS&at_campaign=KARANGA strabane 2022-11-18 10:49:35
ニュース BBC News - Home Autumn Statement: BBC experts on six things you need to know https://www.bbc.co.uk/news/business-63386356?at_medium=RSS&at_campaign=KARANGA budget 2022-11-18 10:41:57
ニュース BBC News - Home What is the minimum wage and how much is it? https://www.bbc.co.uk/news/uk-politics-48445674?at_medium=RSS&at_campaign=KARANGA april 2022-11-18 10:01:02
ニュース Newsweek サメの体内から豚足が見つかる https://www.newsweekjapan.jp/stories/world/2022/11/post-100146.php このオオメジロザメはブタを食べたんだ」【動画】釣ったサメの中から豚足が出てくる瞬間豪ボンド大学のダリル・マクフィー教授は「オオメジロザメのような大きなサメの胃袋からくず肉が見つかっても不思議ではない」と語る。 2022-11-18 19:15:00
IT 週刊アスキー 『Marvel's Spider-Man Miles Morales』PC版が11月19日についに発売! https://weekly.ascii.jp/elem/000/004/113/4113825/ epicgames 2022-11-18 19:45:00
IT 週刊アスキー AVerMedia、ゲームキャプチャー「GC551G2」とコンデンサマイク「AM310」のホワイトバージョンを限定生産で発売 https://weekly.ascii.jp/elem/000/004/113/4113801/ avermedia 2022-11-18 19:30:00
IT 週刊アスキー QNAP、2.5GbE LANポート2基搭載の2ベイNAS「TS-264」 https://weekly.ascii.jp/elem/000/004/113/4113800/ celeron 2022-11-18 19:10:00
海外TECH reddit Qatar Bans Beer Sales at World Cup Stadiums https://www.reddit.com/r/worldnews/comments/yyghiy/qatar_bans_beer_sales_at_world_cup_stadiums/ Qatar Bans Beer Sales at World Cup Stadiums submitted by u za to r worldnews link comments 2022-11-18 10:05:25

コメント

このブログの人気の投稿

投稿時間:2021-06-17 05:05:34 RSSフィード2021-06-17 05:00 分まとめ(1274件)

投稿時間:2021-06-20 02:06:12 RSSフィード2021-06-20 02:00 分まとめ(3871件)

投稿時間:2020-12-01 09:41:49 RSSフィード2020-12-01 09:00 分まとめ(69件)