Benchmarking Google Cloud DataFlow

Google Cloud Dataflow Benchmark
dataflow tops spark benchmark test
Google Cloud Dataflow
ApacheFlink-DataFlow (use Beam Runner)
ApacheSpark-DataFlow (use Beam Runner)
Beam vs Spark Comparison

NER Projects

Named Entity Recognizers are a form of information extraction focusing precisely on named entities in order to classify them into specifically defined categories which may utilize entity linking. Annotation is a fundamental aspect of this classification. Quality measures often incorporate the use of precision, recall and F1 score (harmonic mean). Evaluations are also often compared against a gold standard: a benchmark that is available under reasonable conditions or the most accurate test possible without restrictions which is defined as the ground truth for the absolute state of information. The below highlight a few open source and commercial projects for NER. One can even utilize semantic web in form of a thesaurus server to incorporate SKOS schemes as a way of classification or annotation of terms in form of embedded URIs. One can view further examples from applications of PoolParty or Apache Stanbol.

Stanford Named Entity Recognizer

Illinois Named Entity Tagger

OpenCalais

Lingpipe Alias-I

AlchemyAPI

Refine-NER Extension

GATE

DBPEDIA SpotLight

OpeNER

Other Libraries for custom NER:
OpenNLP
UIMA
CORENLP
SPACY
NLTK
SyntaxNet & TensorFlow
DL4J
Apache Lucene
KEA
FastText

SpeedRead
Knowledge Population
Benchmarking
NER Survey
Google NLP API

Babelfy

NERD

FRED

Dexter

FREME

xLisa
FOX
AIDA

Words and Vectors

Clustering has become an active research area driven through deep learning techniques in deriving vectors of understanding in Natural Language Processing. Word2Vec is a fairly actively used technique for clustering. Its input is a text corpus and its output is a set of feature vectors for words. There are many libraries available that provide implementations for word embeddings including Gensim, DL4J, Spark, and others. The following are some variational areas within the same Word2Vec approach.

Doc2Vec (aka Paragraph2Vec, Sentence2Vec, Text2Vec)
Phrase2Vec
Sequence2Vec

Thought2Vec

GloVe
Concept2Vec
Sense2Vec

Word2Vec Paper
Word2Vec Paper2
Word2Vec Paper3
Google Blog on Word2Vec
Background on Word2Vec

Open Semantic Search

Seems like a new open source project in semantic search, quite useful in the coverage of features that they are trying to achieve. Although, it appears it is still a very new project with much to be implemented. However, tracking it would be still very useful.

Open Semantic Search

Popular BigData and Machine Learning Libraries

Machine Learning libraries and frameworks are constantly evolving. However, there is no harmonization with one tool that fits all solutions.  It seems quite apparent that as more and more libraries evolve the plethora of Machine Learning libraries to choose from will grow to such levels that they will eventually be shunned and refactored towards the cloud in order to utilize greater data processing requirements for scale out. However, certain libraries have a massive following already in industry as examples of some are listed below. Languages like Python, Java, Scala, and C++ are most suited to such contextual work. However, languages like Go are not far behind either. Most of these libraries are directly related to the progress in academic research in the area which can equally provide an indication of what new approaches can be utilized now and what may be possible in the future.

TensorFlow
DL4J
DataFlow
Flink
Spark
Theano
ScikitLearn
GraphLab
Mahout
SpringXD

Game Theory

Game Theory in many respects is the bedrock of advertising and algorithmic trading markets. Combine this with Complex Networks and formal Machine Learning and one has a decisive strategy model. The mathematical models are also applied in Multiagent Systems for studying Argumentation Theory and communication between agents. A Beautiful Mind was a movie that perhaps made Game Theory and the concept of Nash Equilibrium a more mainstream concept. There are endless applications to the field and a few reference sources of further study are provided below.

Graph Theoretic Methods in Multiagent Networks
Multiagent Systems
MAS: Algorithmic, Game-Theoretic, and Logical Foundations

Algorithmic Game Theory

Game Theory: An Introduction

Mathematical Game Theory and Applications
Networks, Crowds, and Markets

Coursera: Game Theory

Coursera: Game Theory II : Advanced Applications

Complex Networks

The rising scale of data and the need for information gain has provided a greater need towards understanding patterns to form knowledgeable insights. In many cases, such patterns can be derived through machine learning and data mining. But, also through studying complex networks that form within contextual data. The below links provide useful sources of study in the science of complex networks.

The Structure and Dynamics of Networks

Networks, Crowds, and Markets

Networks: An Introduction

Understanding Social Networks

Six Degrees: The Science of Connected Age

Coursera: Networked Life

Coursera: Social Network Analysis

Network Science

Network Science Book

Daily Free Packt EBooks

Free Ebooks are available daily from Packt eLearning.

Free Learning