Blockchain developer at MarketPay.io - Theoretical Physicist - R&D
Fintech meetUp about Blockchain, my exposition at 1h46m35s:
Go check out my last developments:
* bloopish.com/ a new Search Engine in town!
* play.google.com/store/apps/details?id=com.proyectos_ip.sheeps An Android Game
**** DATA SCIENCE ****
* Latent Dirichlet Allocation with R language = www.linkedin.com/pulse/latent-dirichlet-allocation-r-language-juan-ignacio-pérez-sacristán
* Are LIGO Gravitational Waves real or just noise? = www.linkedin.com/pulse/ligo-gravitational-waves-real-just-noise-juan-ignacio-pérez-sacristán
* Art with R language = www.linkedin.com/pulse/art-r-language-juan-ignacio-pérez-sacristán
**** SEMANTIC WEB ARCHITECT ****
Getting knowledge from Big Data is today's challenge. Finding out the truth is tomorrow's.
Social Mining 100 million items, developing real-time search engine, extracting triples (subject-predicate-object), correlating entities, weighting Entity-Rank, finding Trends, my own Agglomerative Hierarchical Clustering algorithm ... Testing a disruptive innovation, a paradigm shift: What if we get a fuzzy Knowledge Representation based on Hilbert spaces from Quantum Mechanics? Current formalism (a knowledge database or graph) lacks of a metric specifying whether two knowledges are similar or somehow related.
*** Data Scientist
*** IT Researcher
*** Semantic Web 3.0 Developer
*** Next generation Search Engine Architect
*** Social Network Platform Builder
Following projects such as:
Semantria, AlchemyAPI, OpenCalais, Gnip, gdeltproject.org, Google WikiLinks, ElasticSearch, SphinxSearch, Inbenta, Bitext, ...
- *** Web 3.0 ***: Big Data (Macrodatos), SemWeb, Semantic Map, Ontology Graphs, Topic maps, Disambiguation, Answer engines, NLP, Web scrapping, Keyword mapping, synonym & stemming, TF-IDF, Vector Space Model, Fast and massive Search Engines, Taxonomy (knowledge categorization), Deep Web Crawling, unstructured data, Wikipedia as Corpus/corpora, n-grams (bigrams, trigrams), Clustering & classification, Soundex, Metaphone & Levenshtein metrics, fuzzy search, Social Media identity, Sentiment Analysis, Markov model of text, Information Retrieval, Context Awareness, Text Summarization, Maximum Entropy.
- *** A.I. ***: Named Entity Recognition (NER), Algorithms, Document Parsing and Information Retrieval (IR), Machine Learning (ML) and Natural Language Processing (NLP).
This is an example of extracting visualization and insight from Big Data. From hundreds of thousands of job entries in several spanish marketplaces, a NER process detects topics and their correlation. Using a physics-based algorithm we obtain a map of concepts, the more close the more correlated:
Sorting a set is a very usual process in Computer Science, but here we are trying to sort a set embedded in a two-dimensional space according to a distance metric (criteria) given by the semantic correlation previously found between the items of the set. As this being a step-by-step process, we find problems when too many items are present, because one of them could be blocking others to reach their optimal position. Well-known sorting algorithms as bubble sort or quicksort are not useful to optimize a 2D spatial distribution as in this case.
Quotes from my soul:
* Both current theoretical physics and knowledge representation lack of a solid metaphysical ground, and without it their evolution and success will remain freeze in the future.
* Internet gives me money to live a life and a playground to creativity.
* Against general thinking, I must say that Quantum theory is incomplete and incompletable.
* Intelligence is to achieve the target without checking every possible alternative.
* "What Computers can't do?" and "What Genes can't do?" are two great books. "What Quantum Physics can't do?" is not yet a book, but someone should write it eventually.
And this is how I spend my spare time: