In representation learning, knowledge graph embedding (KGE), also called knowledge representation learning (KRL), or multi-relation learning, is a machine learning task of learning a low-dimensional representation of a knowledge graph's entities and relations while preserving their semantic meaning. Leveraging their embedded representation, knowledge graphs can be used for various applications such as link prediction, triple classification, entity recognition, clustering, and relation extraction. == Definition == A knowledge graph G = { E , R , F } {\displaystyle {\mathcal {G}}=\{E,R,F\}} is a collection of entities E {\displaystyle E} , relations R {\displaystyle R} , and facts F {\displaystyle F} . A fact is a triple ( h , r , t ) ∈ F {\displaystyle (h,r,t)\in F} that denotes a link r ∈ R {\displaystyle r\in R} between the head h ∈ E {\displaystyle h\in E} and the tail t ∈ E {\displaystyle t\in E} of the triple. Another notation that is often used in the literature to represent a triple (or fact) is ⟨ head , relation , tail ⟩ {\displaystyle \langle {\text{head}},{\text{relation}},{\text{tail}}\rangle } . This notation is called the Resource Description Framework (RDF). A knowledge graph represents the knowledge related to a specific domain; leveraging this structured representation, it is possible to infer a piece of new knowledge from it after some refinement steps. However, nowadays, people have to deal with the sparsity of data and the computational inefficiency to use them in a real-world application. The embedding of a knowledge graph is a function that translates each entity and each relation into a vector of a given dimension d {\displaystyle d} , called embedding dimension. It is even possible to embed the entities and relations with different dimensions. The embedding vectors can then be used for other tasks. A knowledge graph embedding is characterized by four aspects: Representation space: The low-dimensional space in which the entities and relations are represented. Scoring function: A measure of the goodness of a triple-embedded representation. Encoding models: The modality in which the embedded representation of the entities and relations interact with each other. Additional information: Any additional information coming from the knowledge graph that can enrich the embedded representation. Usually, an ad hoc scoring function is integrated into the general scoring function for each additional piece of information. == Embedding procedure == All algorithms for creating a knowledge graph embedding follow the same approach. First, the embedding vectors are initialized to random values. Then, they are iteratively optimized using a training set of triples. In each iteration, a batch of size b {\displaystyle b} triples is sampled from the training set, and a triple from it is sampled and corrupted—i.e., a triple that does not represent a true fact in the knowledge graph. The corruption of a triple involves substituting the head or the tail (or both) of the triple with another entity that makes the fact false. The original triple and the corrupted triple are added in the training batch, and then the embeddings are updated, optimizing a scoring function. Iteration stops when a stop condition is reached. Usually, the stop condition depends on the overfitting of the training set. At the end, the learned embeddings should have extracted semantic meaning from the training triples and should correctly predict unseen true facts in the knowledge graph. === Pseudocode === The following is the pseudocode for the general embedding procedure. algorithm Compute entity and relation embeddings input: The training set S = { ( h , r , t ) } {\displaystyle S=\{(h,r,t)\}} , entity set E {\displaystyle E} , relation set R {\displaystyle R} , embedding dimension k {\displaystyle k} output: Entity and relation embeddings initialization: the entities e {\displaystyle e} and relations r {\displaystyle r} embeddings (vectors) are randomly initialized while stop condition do S b a t c h ← s a m p l e ( S , b ) {\displaystyle S_{batch}\leftarrow sample(S,b)} // Sample a batch from the training set for each ( h , r , t ) {\displaystyle (h,r,t)} in S b a t c h {\displaystyle S_{batch}} do ( h ′ , r , t ′ ) ← s a m p l e ( S ′ ) {\displaystyle (h',r,t')\leftarrow sample(S')} // Sample a corrupted fact T b a t c h ← T b a t c h ∪ { ( ( h , r , t ) , ( h ′ , r , t ′ ) ) } {\displaystyle T_{batch}\leftarrow T_{batch}\cup \{((h,r,t),(h',r,t'))\}} end for Update embeddings by minimizing the loss function end while == Performance indicators == These indexes are often used to measure the embedding quality of a model. The simplicity of the indexes makes them very suitable for evaluating the performance of an embedding algorithm even on a large scale. Given Q {\displaystyle {\ce {Q}}} as the set of all ranked predictions of a model, it is possible to define three different performance indexes: Hits@K, MR, and MRR. === Hits@K === Hits@K or in short, H@K, is a performance index that measures the probability to find the correct prediction in the first top K model predictions. Usually, it is used k = 10 {\displaystyle k=10} . Hits@K reflects the accuracy of an embedding model to predict the relation between two given triples correctly. Hits@K = | { q ∈ Q : q < k } | | Q | ∈ [ 0 , 1 ] {\displaystyle ={\frac {|\{q\in Q:q This is a list of software and programming tools for the Ruby programming language, which includes libraries, web frameworks, implementations, tools, and related projects. == Web tools == Capistrano (software) – remote server automation tool Mongrel – Ruby web server Rack – interface between web servers and web applications Ruby on Rails – full-stack web application framework Sinatra – lightweight Ruby web application framework Spree Commerce – e-commerce platform WEBrick – Ruby HTTP server toolkit == Libraries == BioRuby – bioinformatics and computational biology library for Ruby Bogus – Ruby library for creating reliable test doubles with contract verification ERuby – embedded Ruby templating EventMachine – event-driven I/O library Factory Bot – test fixtures library Fat comma – Ruby library for JSON-like hash syntax Geocoder – Ruby library for geocoding and reverse geocoding addresses Haml – HTML templating engine Markaby – HTML generation via Ruby Nokogiri – XML/HTML parsing library RSpec – behavior-driven testing framework for Ruby RubyGems – package manager for Ruby libraries and applications Sass – CSS preprocessor Sidekiq – background job framework for Ruby, used to handle asynchronous tasks. Uconv – Unicode text conversion library Watir – web application testing framework == Ruby implementations == HotRuby – Ruby interpreter implemented in JavaScript, enabling Ruby code to run in web browsers. IronRuby – Ruby for .NET platform JRuby – Ruby on the Java Virtual Machine MacRuby – Ruby implementation for macOS Mod ruby – Apache module that embeds the Ruby interpreter to improve performance of Ruby web applications Mruby – lightweight Ruby interpreter Rubinius – alternative Ruby implementation, based loosely on the Smalltalk-80 Blue Book design. Ruby MRI – the standard Ruby interpreter YARV – "Yet Another Ruby VM," the bytecode interpreter used in modern Ruby implementations == Tools == Homebrew – package manager for macOS and Linux written in Ruby Pry – interactive Ruby shell Rake – build and task management Ruby Version Manager – environment manager RubyCocoa – bridge between Ruby and Cocoa RubyForge – project hosting site RubyMotion – for iOS/macOS development RubySpec – language specification tests == Integrated Development Environments == Aptana Studio — integrated RadRails plugin for Ruby on Rails development Eclipse DLTK Ruby Plugin — Ruby development plugin for Eclipse Eric — open-source Python-based IDE with Ruby support Komodo IDE — commercial cross-platform IDE with Ruby support RubyMine — commercial IDE for Ruby and Rails by JetBrains SlickEdit — commercial cross-platform IDE with Ruby support == List of websites using Ruby on Rails == Airbnb Basecamp Diaspora – decentralized social network application built with Ruby on Rails Discourse – open-source discussion platform built with Ruby on Rails Fiverr GitHub Hulu Shopify SoundCloud Twitch Zendesk IBM 37xx (or 37x5) is a family of IBM Systems Network Architecture (SNA) programmable front-end processors used mainly in mainframe environments. All members of the family ran one of three IBM-supplied programs. Emulation Program (EP) mimicked the operation of the older IBM 270x non-programmable controllers. Network Control Program (NCP) supported Systems Network Architecture devices. Partitioned Emulation Program (PEP) combined the functions of the two. == Models == === 370x series === 3705 — the oldest of the family, introduced in 1972 to replace the non-programmable IBM 270x family. The 3705 could control up to 352 communications lines. 3704 was a smaller version, introduced in 1973. It supported up to 32 lines. === 371x === The 3710 communications controller was introduced in 1984. === 372x series === The 3725 and the 3720 systems were announced in 1983. The 3725 replaced the hardware line scanners used on previous 370x machines with multiple microcoded processors. The 3725 was a large-scale node and front end processor. The 3720 was a smaller version of the 3725, which was sometimes used as a remote concentrator. The 3726 was an expansion unit for the 3725. With the expansion unit, the 3725 could support up to 256 lines at data rates up to 256 kbit/s, and connect to up to eight mainframe channels. Marketing of the 372x machines was discontinued in 1989. IBM discontinued support for the 3705, 3720, 3725 in 1999. === 374x series === The 3745, announced in 1988, provides up to eight T1 circuits. At the time of the announcement, IBM was estimated to have nearly 85% of the over US$825 million market for communications controllers over rivals such as NCR Comten and Amdahl Corporation. The 3745 is no longer marketed, but still supported and used. The 3746 "Nways Controller" model 900, unveiled in 1992, was an expansion unit for the 3745 supporting additional Token Ring and ESCON connections. A stand-alone model 950 appeared in 1995. == Successors == IBM no longer manufactures 37xx processors. The last models, the 3745/46, were withdrawn from marketing in 2002. Replacement software products are Communications Controller for Linux on System z and Enterprise Extender. == Clones == Several companies produced clones of 37xx controllers, including NCR COMTEN and Amdahl Corporation. In database systems, atomicity (; from Ancient Greek: ἄτομος, romanized: átomos, lit. 'undividable') is the property of a database transaction consisting of an indivisible and irreducible series of database operations such that either all occur, or none occur. It is one of the ACID transaction properties: Atomicity, Consistency, Isolation, Durability. A guarantee of atomicity prevents partial database updates from occurring, because they can cause greater problems than rejecting the whole series outright. As a consequence, an atomic transaction cannot be observed to be in progress by another database client: at one moment in time, it has not yet happened, and at the next it has already occurred in whole (or nothing happened if the transaction was cancelled in progress). An example of transaction atomicity could be a digital monetary transfer from bank account A to account B. It consists of two operations, debiting the money from account A and crediting it to account B. Performing both of these operations inside of an atomic transaction ensures that the database remains in a consistent state, if either operation fails there will not be any unaccountable credits or debits affecting either account. The same term is also used in the definition of First normal form in database systems, where it instead refers to the concept that the values for fields may not consist of multiple smaller values to be decomposed, such as a string into which multiple names, numbers, dates, or other types may be packed. == Orthogonality == Atomicity does not behave completely orthogonally with regard to the other ACID properties of transactions. For example, isolation relies on atomicity to roll back the enclosing transaction in the event of an isolation violation such as a deadlock; consistency also relies on atomicity to roll back the enclosing transaction in the event of a consistency violation by an illegal transaction. As a result of this, a failure to detect a violation and roll back the enclosing transaction may cause an isolation or consistency failure. == Implementation == Typically, systems implement Atomicity by providing some mechanism to indicate which transactions have started and which finished; or by keeping a copy of the data before any changes occurred (Read-copy-update). Several filesystems have developed methods for avoiding the need to keep multiple copies of data, using journaling (see journaling file system). Databases usually implement this using some form of logging/journaling to track changes. The system synchronizes the logs (often the metadata) as necessary after changes have successfully taken place. Afterwards, crash recovery ignores incomplete entries. Although implementations vary depending on factors such as concurrency issues, the principle of atomicity – i.e. complete success or complete failure – remain. Ultimately, any application-level implementation relies on operating-system functionality. At the file-system level, POSIX-compliant systems provide system calls such as open(2) and flock(2) that allow applications to atomically open or lock a file. At the process level, POSIX Threads provide adequate synchronization primitives. The hardware level requires atomic operations such as Test-and-set, Fetch-and-add, Compare-and-swap, or Load-Link/Store-Conditional, together with memory barriers. Portable operating systems cannot simply block interrupts to implement synchronization, since hardware that lacks concurrent execution such as hyper-threading or multi-processing is now extremely rare. In distributed and sharded databases, atomicity is complicated by network latency and the potential for partial failures. While traditional distributed systems often employ locking protocols (like 2PC) to ensure cross-shard atomicity, these can introduce performance bottlenecks. Recent research into distributed ledger consensus suggests alternative models, such as "braided synchronization". This technique, utilized in protocols like Cerberus, intertwines the consensus phases of multiple shards to enforce atomic guarantees without a global ordering of all transactions. The CARE Principles for Indigenous Data Governance are a set of principles intended to guide open data projects in engaging Indigenous Peoples rights and interests. CARE was created in 2019 by the International Indigenous Data Sovereignty Interest Group, a group that is a part of the Research Data Alliance. It outlines collective rights related to open data in the context of the United Nations Declaration on the Rights of Indigenous Peoples and Indigenous data sovereignty. CARE is an acronym which stands for Collective Benefit, Authority to Control, Responsibility, Ethics. The CARE Principles are 'people and purpose-oriented, reflecting the crucial role of data in advancing Indigenous innovation and self-determination', and intended as a complement to the data-oriented perspective of other standards such as FAIR data (findable, accessible, interoperable, reusable). The CARE principles have been embedded into the Beta version of Standardised Data on Initiatives (STARDIT). CARE principles were the basis of a submission to the UN's Global Digital Compact. LTX is a family of open source artificial intelligence video foundation models developed by Lightricks, and first released in November 2024. The latest models, LTX-2, create videos based on user prompts. They were preceded by LTX Video, which was released in 2024 as the company's first text-to-video model. LTX-2 is part of the LTX family of video generation models, which form the core technology, alongside LTX Studio, of the LTX ecosystem. == History == === Origins: LTX Video (2024–2025) === In November 2024 Lightricks publicly released its first text-to-video model, LTX Video. It was a 2-billion parameter model, available as open source. In May 2025 Lightricks launched LTXV-13b, a version with 13-billion parameters. Two months later, the model broke the 60 second barrier for generated video. === Release of LTX-2 (2025) === In October 2025 Lightricks announced its latest model, and renamed it LTX-2. The model was described as capable of generating synchronized audio and video at native 4K resolution and up to 50 frames per second (fps), using a variety of conditions and prompts, including text-to-video and image-to-video. Google highlighted the fact that LTX-2 was trained on its infrastructure, and saying it was "The first open source AI video generation model, powered by Google Cloud". Upon its release it was ranked in the top-3 models for image-to-video creation by Artificial Analysis, behind Kling 3.5 by Kling AI and Veo 3.1 by Google. Its text-to-image option was ranked 7th. In addition to its open-source release, Lightricks offers API access to LTX-2, allowing developers to generate videos from text and image prompts through a hosted service without running the model locally. === Open Source Release (2026) === In January 2026, Lightricks officially released the full open-source version of LTX-2, making the model’s complete codebase, weights, and associated tooling publicly available. In March 2026 the company released LTX-2.3, which was accompanied by a desktop video editor enabling the entire model to run locally on consumer hardware. == Technical features == === Advancements over LTX Video === LTX-2 builds upon the LTX Video architecture with several major improvements: Unified audio-video generation producing synchronized dialogue, ambience, and motion Native 4K rendering 50-fps output for cinematic motion Three operational modes (Fast, Pro, Ultra) More efficient diffusion pipelines enabling high fidelity on consumer GPUs === Core capabilities === Text-to-video generation Image-to-video generation Multimodal audiovisual synthesis High-resolution spatial and temporal coherence Configurable quality/performance settings Open-source distribution of weights and datasets == Reception == Initial reception to LTX-2 was broadly positive, with several technology and media outlets highlighting its open-source approach and multimodal capabilities. Open Source For You described LTX-2 as “one of the first AI video systems to combine 4K output, synchronized audio, and an open model release,” noting that it positioned Lightricks as a significant competitor to proprietary systems such as OpenAI's Sora and Google's Veo. IEA Green said that the model “could rewrite the AI filmmaking game,” emphasizing that its 50-fps rendering and unified audio-video generation made it suitable for professional studios and independent creators alike. AI News characterized LTX-2 as a “major step forward in the democratization of cinematic-quality video generation,” praising its consumer-grade hardware efficiency and multi-tier generation modes, while also noting ongoing challenges in long-form temporal stability. FinancialContent reported strong interest among creative agencies, attributing the attention to Lightricks’ decision to release model weights and datasets, which reviewers said enabled “a level of transparency not typically seen in commercial AI video models.” === Benchmarks and rankings === Upon release, LTX-2 ranked third for image-to-video creation in the Artificial Analysis benchmark, behind Kling 3.5 and Veo 3.1, while its text-to-video option ranked seventh. As of early 2026, it was the highest-ranked open-source model in the benchmark. === Limitations === Some early reviewers also pointed out quality limitations. The Ray3 technical review noted occasional inconsistencies in lip-sync and motion tracking during long scenes, though it stated these were “in line with the challenges faced by all current AI video diffusion models” and expected to improve with continued iteration. Like other diffusion-based video generators, LTX-2 can produce artifacts in complex multi-person scenes and may struggle with precise text rendering within generated video. Data marketplace is an online platform for sharing and consuming data in the form of data assets or data products. Part of the data management stack, it aims to bring together data producers and data consumers (including business users and AI) in a single space, with the objective of increasing access to understandable, high-quality data. Included within its Data Marketplaces and Exchange (DME) category by Gartner, data marketplaces can provide data internally within an organization, externally with partners, or as open data. == Concept == Digitization has dramatically increased data volumes within organizations, with IDC predicting that by 2025 the world will contain 175 zettabytes of data. This has created a need to both manage this data and provide access to it to enable business intelligence and data analysis. However, data is often scattered within multiple systems (such as data warehouses and data lakes), and is in formats that are only understandable by technical experts, such as data scientists. According to IDC, 81% of IT leaders cite data silos as a major barrier to digital transformation. This means that data is not freely available to business users or external audiences such as partners or citizens, limiting its value, and holding back AI deployments. Data marketplaces solve this issue, providing seamless, self-service access to high-quality data in an understandable, secure and auditable manner. They break down data silos, reduce friction in data access, and enable a broader range of users, including non-technical profiles, to find, understand, and consume data autonomously. Data assets on the marketplace can be raw data, data visualizations or data products. Data marketplaces combine data management functions such as data governance with the user-friendly experience offered by e-commerce marketplaces in order to increase the usage of data. These include features such as powerful search engines, feedback, ratings, subscriptions and product description sheets. According to Gartner, data marketplaces provide infrastructure, transactional capabilities, and services for both consumers and providers of data assets. == History and timeline == Data marketplaces have evolved since they first emerged in terms of both their scope and usage. === 2000s === With the rise of the internet, data brokers began collecting, aggregating, distributing and selling personal, financial and marketing data to third parties online. Data marketplaces were deployed to monetize this data, making it discoverable and accessible to users, either through subscriptions or one-off purchases. At the same time, regulations, such as the US Open Government Initiative of 2009 and others around the world mandated greater transparency and data sharing with the public. Data sharing portals were created by public and government bodies to make this information available through self-service to all users. === 2010s === Due to the growth of big data and cloud platforms, cloud-based data exchange platforms emerged. These were offered by major infrastructure providers, and included Amazon Web Services (AWS) Data Exchange, Snowflake Data Marketplace, and the Google Cloud Platform. These platforms moved beyond simple data brokerage or open data by providing structured, catalogued data sharing between organizations. === 2020s === Driven by a need to increase internal data sharing with both business users and AI, organizations are now looking to adopt internal data marketplaces. These aim to democratize data consumption by providing seamless access for all employees and AI to trusted data, including data products, through an intuitive, e-commerce style experience. According to Gartner analyst Richa Jha, "by providing a single, governed platform for discovering, sharing, and scaling data products, data marketplaces drive productivity, collaboration, and ROI across the enterprise." == Data marketplaces within the overall data architecture == Data marketplaces provide a consumption and collaboration layer for data. That means they complement and integrate with other parts of the overall data architecture, including: === Data warehouses and data lakes === Data marketplaces connect to data sources, such as data warehouses or data lakes, to provide intuitive access to the data stored within them, enabling data to be shared and distributed to non-technical audiences. Access can be direct, with data and data products stored within the data marketplace or virtualized. === Data catalog === A data catalog provides a technical inventory of an organization's data estate. It collects technical information on all available data assets within an organization, based on metadata descriptions. This ensures traceability, and supports compliance and governance requirements. Unlike a data marketplace, a data catalog does not provide access to data, and is designed to be used by data professionals, rather than the business. This means it lacks an intuitive, understandable interface and is consequently not easily accessible by business users. === Data mesh === Data mesh is an architecture and framework for data management, first defined by Zhamak Dehghani in 2019. It aims to decentralize data ownership to delegate responsibility, empowering teams and focusing on delivering data to users in the form of self-service data products. The data marketplace is a central pillar of data mesh, providing intuitive access to these data products, and creating a collaboration space for data owners and data consumers. === Data product === Data products are high-value, consumable data assets that package high-quality data and associated tools to enable seamless usage by business users at scale. First defined by McKinsey in 2022, they have an identified owner, a service level agreement (SLA), and a reusability logic. == Core components of a data marketplace == A data marketplace typically includes specific core components: === E-commerce style interface === An e-commerce style experience that engages non-technical users, minimizes the need for training and builds confidence and trust in data. Look and feel should be customizable to incorporate corporate design guidelines to ensure consistency with other organizational applications. === Built-in data catalog === As in a standalone data catalog, this indexes all available data, based on metadata that includes type, source, owner, freshness, and quality level. === Discovery and search engine === This enables users to search, filter, explore and discover available data intuitively. As in an e-commerce marketplace, it should be intelligent, and provide relevant results based on natural language queries. === Access control and security management === Data marketplaces will contain data that needs to be protected under regulations such as the General Data Protection Regulation (GDPR) in Europe, the California Consumer Privacy Act (CCPA) in the United States, and sector-specific frameworks in industries such as finance and healthcare. To ensure both security and compliance while maximizing data consumption, the data marketplace should include granular access management and a full audit trail. === Semantic layer and business glossary === Different parts of the business are likely to use different terms to describe data. This leads to inconsistencies and an inability to share data across systems and teams. The semantic layer and business glossary standardize a shared vocabulary and common definitions of business indicators and concepts, providing a single language for data across the business and for AI agents. === Data governance mechanisms === These enforce corporate data governance policies, ensuring data traceability through data lineage, quality certification, usage monitoring, and continuous improvement through user feedback loops. === Collaboration features === As on an e-commerce website, a data marketplace should provide collaboration features that bring together data users and data owners. This includes the ability to rate data products, share use cases, and provide feedback to data owners, creating a community around data and supporting a data-driven culture. == Types of data marketplace == While they share the same underlying technology, data marketplaces can be deployed in three broad ways: === Internal data marketplaces === These bring together data from across an organization and make it available via self-service to employees from across the business. They aim to widen access to data and consequently to improve decision-making and reporting, increase performance and maximize efficiency. === Ecosystem data marketplaces === These extend sharing beyond a single organization, enabling multiple partners (public institutions, industry players, research bodies) to share and consume data within a governed framework. Data can be provided by all parties or simply by one organization and consumed by others. Ecosystem data marketplaces are particularly relevant in List of Ruby software and tools
IBM 37xx
Atomicity (database systems)
CARE Principles for Indigenous Data Governance
LTX (text-to-video model)
Data marketplace