contador Saltar al contenido

Advances in cognitive computing, the computer that (almost) thinks

The cognitive compuntig takes another step forward and the computer is getting closer to what everyone thinks is an insurmountable barrier: thinking.

Watson "width =" 2880 "height =" 1920 "class =" alignleft size-full wp-image-80398 "srcset =" jpg 2880w, 300w, Watson-1024x682.jpg 1024w "sizes =" (max-width: 2880px) 100vw, 2880px "/> Today, of course, the computer does not only perform data analysis, at least not Watson. IBM states, in fact, that it has made significant progress in the cognitive computing capabilities of Watson, overcoming the phase of data analysis and research of known answers to move to the exploration of the complexity and connections of enormous amounts of data.</p><div class=

This ability almost to even think of a service available immediately and provided as a service in the cloud, Watson Discovery Advisor designed to accelerate the discoveries by the research teams, because it reduces the time needed to test the hypotheses and formulate conclusions, from months to days and from days to a few hours, bringing new levels of speed and precision to research and development. Thanks to Watson's ability to grasp the nuances of natural language, Watson Discovery Advisor can understand the language of science, for example how compounds interact chemists, thus representing an extraordinarily powerful tool for researchers in biosciences and other fields.

Researchers and scientists from leading university, pharmaceutical and research centers have started using IBM's new Watson Discovery Advisor to analyze and test hypotheses using the millions of scientific documents available in public databases. Today a new study is published every 30 seconds, which is equivalent to more than one million studies a year (source: CiteSeerx). According to the National Institutes for Health, a researcher generally reads about 23 scientific documents a month, equivalent to nearly 300 a year, making it humanly impossible to keep up with the ever-increasing volume of available scientific material. In 2013, the top 1000 Research companies and Development spent more than $ 600 billion a year just on Research. Progress can be slow, it takes an average of 10 to 15 years for a promising pharmaceutical treatment to move from the initial research phase to practice (source: Pharmaceutical Research and Manufacturers of America). Thanks to Watson Discovery Advisor, researchers can discover new relationships and identify unexpected patterns in data, with the possibility of significantly increasing and accelerating discovery processes. We are entering an extraordinary era in which the discovery will be driven by data, explains Mike Rhodin , senior vice president, IBM Watson Group. Today's announcement represents a natural extension of Watson's cognitive computing intelligence and a powerful tool for researchers, developers and industry experts that will help to increase the results of the investments made by organizations in R&D, leading to significant revolutionary discoveries. Leading organizations in biosciences are using Watson Discovery Advisor to advance their research projects, including Baylor College of Medicine, Johnson & Johnson and the New York Genome Center.

Impossible questionsIn a peer-reviewed, retrospective study published this week by Baylor College of Medicine and IBM, scientists demonstrated a possible new path to generate scientific questions that could be useful in the long term for the development of new effective treatments for diseases . Using Baylor's Knowledge Integration Toolkit (KnIT) based on Watson's technology, within weeks, biologists and data scientists have accurately identified p53-changing proteins, an important protein related to many types of cancer: this result will improve efficacy of drugs and other treatments. A company whose realization, without the cognitive functionalities of Watson, would have required researchers to work years, in fact Watson analyzed 70,000 scientific articles on p53, to predict which proteins activate or deactivate it. This automated analysis led Baylor's cancer researchers to identify six potential proteins that will be the subject of new research. These results are relevant, considering that in the last 30 years scientists have discovered on average a similar target protein a year. On average a scientist could read, at most, between one and five research articles a day, comments Dr. Olivier Lichtarge, researcher and professor of molecular and human genetics, biochemistry and molecular biology at Baylor College of Medicine. More than 70,000 published studies have been published on p53. Even reading five a day, it could take me almost 38 years to fully understand all the research already available, Watson accelerates the speed and quality of revolutionary discoveries.