NGT

Neighborhood Graph and Tree for Indexing High-dimensional Data

Home / Installation / Command / License / Publications / About Us / 日本語

NGT provides commands and a library for performing high-speed approximate nearest neighbor searches against a large volume of data (several million to several 10 million items of data) in high dimensional vector data space (several ten to several thousand dimensions).

News

Python NGT can be installed via pip from PyPI (1/17/2019)
NGTQ(NGT with Quantization) is now available (12/14/2018 : v1.5.0)
ONNG is now available. (08/08/2018 : v1.4.0)

Installation

Downloads

Releases

Build

On Linux

  $ unzip NGT-x.x.x.zip
  $ cd NGT-x.x.x
  $ mkdir build
  $ cd build 
  $ cmake ..
  $ make 
  $ make install
  $ ldconfig /usr/local/lib

On Mac using homebrew

  $ /usr/bin/ruby -e "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/master/install)"
  $ brew install cmake
  $ brew install gcc
  $ ln -s ./gcc-8 /usr/local/bin/gcc
  $ ln -s ./g++-8 /usr/local/bin/g++
  $ export CXX=g++
  $ export CC=gcc
  $ unzip NGT-x.x.x.zip
  $ cd NGT-x.x.x
  $ mkdir build
  $ cd build 
  $ cmake ..
  $ make 
  $ make install

Shared memory use

The index can be placed in shared memory. Using shared memory can reduce the amount of memory needed when multiple processes are using the same index. It can also improve the boot-up speed of an index for a large volume of registration data. Since changes become necessary at build time, please add the following parameter when executing "cmake" in order to use shared memory.

  $ cmake -DNGT_SHARED_MEMORY_ALLOCATOR=ON ..

Note: Since there is no lock function, the index should be used only for reference when multiple processes are using the same index.

Utilities

Command : ngt, ngtq
Server : ngtd

Supported Programming Languages

License

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this software except in compliance with the License. You may obtain a copy of the License at

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.

Yahoo Japan Corporation has aquired several patents on the technologies utilized in this software. However, the patent rights shall not be exercised under Apache License Version 2.0, only when the patented techniques are used with this software.

Contributor License Agreement

This project requires contributors to accept the terms in the Contributor License Agreement (CLA).

Please note that contributors to the NGT repository on GitHub (https://github.com/yahoojapan/NGT) shall be deemed to have accepted the CLA without individual written agreements.

Publications

ONNG

Iwasaki, M., Miyazaki, D.: Optimization of Indexing Based on k-Nearest Neighbor Graph for Proximity. arXiv:1810.07355 [cs] (2018). (pdf)

PANNG

Iwasaki, M.: Pruned Bi-directed K-nearest Neighbor Graph for Proximity Search. Proc. of SISAP2016 (2016) 20-33. (pdf)
Sugawara, K., Kobayashi, H. and Iwasaki, M.: On Approximately Searching for Similar Word Embeddings. Proc. of ACL2016 (2016) 2265-2275. (pdf)

ANNGT

Iwasaki, M.: Applying a Graph-Structured Index to Product Image Search (in Japanese). IIEEJ Journal 42(5) (2013) 633-641. (pdf)
Iwasaki, M.: Proximity search using approximate k nearest neighbor graph with a tree structured index (in Japanese). IPSJ Journal 52(2) (2011) 817-828. (pdf)

ANNG

Iwasaki, M.: Proximity search in metric spaces using approximate k nearest neigh-bor graph (in Japanese). IPSJ Trans. on Database 3(1) (2010) 18-28. (pdf)

dbaranchuk/NGT

NGT