Machine learning

Retrieval-augmented generation in which a language model answers from a knowledge graph built over the source texts, rather than from isolated text chunks.

Framework proposed by Creative Commons in June 2025 for dataset holders to state, in machine-readable form, how they want their content reused by AI systems; partly legal and partly a social expectation.

Data too large, fast-moving or varied for ordinary databases, and the methods for drawing value from it by analysis at scale; the term spread around 2010 as companies mined logs, clicks and sensor data.

Neural language model trained on very large text collections; since the early 2020s the main way software works with unstructured text.

AI that combines neural networks with symbolic knowledge such as rules and knowledge graphs; it drew new attention in the 2020s as a way to ground large language models.

Technique in which a language model first retrieves relevant documents or data and then generates its answer from them.

Halevy, Norvig and Pereira's 2009 essay arguing that simple models trained on very large amounts of web data beat elaborate hand-built models, and that meaning is better learned from use at scale than specified in advance.

2012 NIPS paper by Alex Krizhevsky, Ilya Sutskever and Geoffrey Hinton whose deep network, trained on GPUs, won the ImageNet image-recognition contest by a wide margin and turned machine learning toward deep learning.

2013 NIPS paper introducing TransE, which represents each entity in a knowledge graph as a vector and each relation as a translation between vectors, so that missing facts can be predicted; the starting point of knowledge-graph embeddings.

2013 paper by Tomas Mikolov, Kai Chen, Greg Corrado and Jeffrey Dean introducing word2vec: vectors learned from which words occur together, in which relations such as king to queen show up as directions.

2014 KDD paper by Xin Luna Dong and eight Google colleagues describing a knowledge base built by machine: facts extracted from text, tables and page structure across the Web, combined with existing knowledge bases and given probabilities of being true.

2015 IEEE Symposium on Security and Privacy paper by Yinzhi Cao and Junfeng Yang that named machine unlearning: removing the effect of particular training data from a learned model without retraining it from scratch.

2017 paper by Ashish Vaswani and seven Google colleagues introducing the transformer, a network built on attention that became the basis of large language models.

2018 paper by Jacob Devlin, Ming-Wei Chang, Kenton Lee and Kristina Toutanova: a transformer pre-trained on large amounts of text that set new results on many language tasks; Google applied it to Search in 2019.

Survey proposing three ways to combine language models and knowledge graphs: KG-enhanced LLMs, LLM-augmented KGs, and the two working together.

Microsoft Research paper by Darren Edge and nine co-authors introducing the GraphRAG method.

2024 paper by Liu, Semnani, Triedman, Xu, Zhao and Lam with a dataset of real questions from Wikidata's 'Request a Query' forum, where users ask others to write SPARQL for them, and a language-model agent that writes such queries the way an expert would.

October 2026 paper by the W3C Strategy Team, prepared by Philippe Le Hégaret, on how the Web's standards must change as AI agents act for users: agents prefer text and structured data to HTML, JSON-LD, RDF, DCAT and ODRL become more valuable as inputs and guardrails for agents, and Verifiable Credentials and DIDs can identify agents and whom they act for.

Open protocol introduced by Anthropic in November 2024 for connecting AI models to external tools and data sources; donated to the Linux Foundation's Agentic AI Foundation in December 2025.

Lexical database of English begun in 1985 at Princeton's Cognitive Science Laboratory under George A. Miller: nouns, verbs and adjectives grouped into sets of synonyms linked by relations such as broader and part-of. It became a shared vocabulary for language processing and for knowledge bases such as YAGO.

American AI company that introduced the Model Context Protocol in 2024.

Research institute founded by Paul Allen; its first grant helped fund Wikidata's initial development in 2012.

American computer scientist, director of research at Google and co-author of 'Artificial Intelligence: A Modern Approach'.

Machine-learning researcher, co-author of the TransE paper (2013).

Machine learning researcher, first author of the 2012 AlexNet paper.

Machine-learning researcher, lead author of the TransE paper (2013).

Machine learning researcher, first author of 'Attention Is All You Need'.

Co-author of 'Unifying Large Language Models and Knowledge Graphs: A Roadmap' (2023).

Microsoft Research scientist and lead author of the 2024 GraphRAG paper.

Anthropic engineer who co-created the Model Context Protocol.

Computer scientist in natural-language processing, research director at Google; co-author of 'The Unreasonable Effectiveness of Data' (2009).

British-Canadian computer scientist whose work on neural networks underlies deep learning; co-author of AlexNet.

Google machine learning scientist, co-author of the 2013 word2vec paper.

Co-author of the SPINACH paper (2024) on question answering over Wikidata.

AI researcher, co-author of AlexNet and later a co-founder of OpenAI.

Co-author of the SPINACH paper (2024) on question answering over Wikidata.

Researcher at Google, first author of BERT.

Machine-learning researcher, co-author of the TransE paper (2013).

American computer scientist at Google who co-designed MapReduce and co-wrote word2vec.

Co-author of the SPINACH paper (2024) on question answering over Wikidata.

Co-author of 'Unifying Large Language Models and Knowledge Graphs: A Roadmap' (2023).

Columbia University computer scientist, co-author of the 2015 paper that named machine unlearning.

Anthropic engineer who co-created the Model Context Protocol.

Google engineer, co-author of the 2013 word2vec paper.

Natural language processing researcher at Google, co-author of BERT.

Natural language processing researcher at Google, co-author of BERT.

Co-author of 'Unifying Large Language Models and Knowledge Graphs: A Roadmap' (2023).

Researcher at Google, co-author of BERT.

American computer scientist at Stanford, known for compilers and later for conversational agents grounded in knowledge bases; co-author of the SPINACH paper (2024).

Machine-learning researcher, co-author of the TransE paper (2013).

Machine-learning researcher at Google, co-author of the TransE paper (2013).

Stanford researcher and lead author of the SPINACH paper (2024) on language models writing SPARQL for Wikidata.

Co-author of 'Unifying Large Language Models and Knowledge Graphs: A Roadmap' (2023).

Stanford researcher and co-author of the SPINACH paper (2024).

Czech computer scientist who led the word2vec work at Google.

Chinese-American computer scientist who led Google's Knowledge Vault and later Amazon's product knowledge graph.

Co-author of 'Unifying Large Language Models and Knowledge Graphs: A Roadmap' (2023).

Computer security researcher, first author of the 2015 paper that named machine unlearning.

Co-author of 'Unifying Large Language Models and Knowledge Graphs: A Roadmap' (2023).