Artificial intelligence and data

Sentence-BERT

Sentence-BERT (SBERT) is a modification of BERT using siamese and triplet networks to produce fixed-size sentence embeddings, introduced by Nils Reimers and Iryna Gurevych at EMNLP 2019 and released with the sentence-transformers framework.

SentencePiece

SentencePiece is an open-source, language-independent subword tokenizer and detokenizer released by Google in 2018, which trains directly from raw sentences and is used by language models such as T5 and Llama.

Sentiment analysis

Sentiment analysis, also called opinion mining, is the natural language processing task of identifying and classifying opinions, emotions, and attitudes in text by polarity such as positive, neutral, or negative.

Oct 26, 2026

Separation of storage and compute

Separation of storage and compute: cloud data architecture, also called decoupling or disaggregation, where query engines and object storage scale independently. Pioneered by BigQuery (2012) and Snowflake (2014); core pattern of lakehouses.

Seq2seq model

A seq2seq (sequence-to-sequence) model is a neural network architecture that maps variable-length input sequences to variable-length outputs, using an encoder and decoder, applied in machine translation and speech recognition.

Sequence mining

Sequence mining, also known as sequential pattern mining, is a data mining method that finds frequently occurring subsequences in ordered data such as customer transactions, clickstreams, and biological sequences.

Sequence parallelism

Sequence parallelism (SP) is a distributed-training technique that splits a transformer's sequence dimension across multiple devices, letting models train on very long sequences such as the 1-million-token contexts demonstrated by DeepSpeed-Ulysses.

Sequence-to-sequence learning

Sequence-to-sequence learning (seq2seq) is a machine learning approach in which an encoder and decoder neural network map input sequences to output sequences of different lengths.

Sequential minimal optimization

Sequential minimal optimization (SMO) is an algorithm introduced by John Platt in 1998 for training support vector machines by solving small two-multiplier quadratic programming subproblems analytically.

Sequential pattern mining

Sequential pattern mining is a data mining method that discovers frequent ordered sequences of events in sequence databases such as customer purchase histories and event logs.

Sequential recommendation with deep learning

Sequential recommendation with deep learning predicts the next item a user will interact with from their ordered, timestamped interaction history; self-attention models such as SASRec (2018) dominate benchmarks.

Set (abstract data type)

A set, also known as a set data structure, is an abstract data type that stores unique elements with no duplicates, supporting add, remove, membership, union, intersection, and difference.

SHA-2

SHA-2 (Secure Hash Algorithm 2) is a family of cryptographic hash functions specified by NIST in 2002 to replace SHA-1, with members from SHA-224 to SHA-512.

SHA-3

SHA-3, also known as Keccak, is a family of cryptographic hash functions standardized by NIST as FIPS 202, based on the Keccak algorithm that won its 2012 competition.

Shadow detection

Shadow detection is the computer vision task of locating shadow regions in images or video. It improves object detection, tracking, and segmentation, which shadows otherwise degrade.

SHAP

SHAP (SHapley Additive exPlanations) is a feature attribution method that explains individual machine-learning predictions by assigning each feature a Shapley value from cooperative game theory.

Shape analysis (computational geometry)

Shape analysis is the branch of computational geometry that compares and models geometric objects after removing location, scale, and rotation, producing shape distances, correspondences, and mean shapes.

Shape context

Shape context is a shape descriptor in computer vision that represents each point on a shape by a histogram of the other points' relative positions, used for shape matching and object recognition.

Shape from shading

Shape from shading is a computer vision problem: recovering the 3D shape of a surface from the shading variation in a single image, founded by Horn in the 1970s.

ShapeNet

ShapeNet is a large, richly annotated dataset of 3D CAD models for computer vision, graphics, and robotics research, introduced in 2015 by Stanford, Princeton, and TTIC researchers.

Shard (database architecture)

A shard is a horizontal partition of a database: data is split into independent pieces distributed across multiple nodes, letting storage and workload scale beyond one server.

Sharpness-aware minimization

Sharpness-aware minimization (SAM) is a training procedure that perturbs model weights toward higher loss before each descent step, favoring flat minima; introduced by Foret and colleagues in 2020.

Shear stress transport model

The shear stress transport model is a two-equation turbulence model used in computational fluid dynamics, introduced by F. R. Menter in 1994 to predict boundary layers and flow separation.

Shell method (program verification)

The shell method is a refinement-based proof technique for interactive theorem proving in which a generic correctness theorem is proved once and instantiated with concrete data structures.

Shellsort

Shellsort is an in-place comparison sorting algorithm that improves on insertion sort by sorting elements a shrinking gap apart, published by Donald Shell in 1959.

Shenzhen Institute of Artificial Intelligence and Robotics for Society

The Shenzhen Institute of Artificial Intelligence and Robotics for Society (AIRS) is a research institute in Shenzhen, China, established in 2019 on the basis of CUHK-Shenzhen.

Oct 26, 2026

Shift-left data processing

Shift-left data processing moves data quality checks, governance, and processing to the point of data creation instead of downstream fixes. Also called data shift left; built on data contracts and early classification; the 1-10-100 cost rule and producer ownership distinguish it from pipeline-then-fix approaches.