Six Years of Rossi’s Thoughts Visualized (Matt Sevrens)

The following post has been submitted by Matt Sevrens

The Mind of Andrea Rossi Visualized Over Time

With MFMP’s recent discoveries regarding the possibility of gamma emissions and lead being used to thermalize these emissions, what’s most striking to me is that most if not all of this information is and has been on the Journal of Nuclear Physics in plain sight for years.

What I realize is that despite the vast reservoir of knowledge there, the majority of LENR followers don’t have the tools to easily search through Rossi’s thoughts to find patterns and useful pieces of information that could assist with replications. Therefore, I’ve put together the following interactive visualization based on six years of content available on the JoNP:

http://demos.mattsev.com/rossiStream/

Much like a stream, thought ebbs and flows over time. Some days the mind is active and flowing with ideas, on others it is but a trickle. Hence I find a good way to visualize thought is a stream graph. In this particular visualization, I have cut up Rossi’s thoughts into words and made counts of the various words that Rossi has used over the last six years. Furthermore I have enabled the capacity to load and read any thought containing any topic Rossi has discussed. Simply select the words you’re interested in the dropdown, click and drag the upper section of the graph to select a time period and then click load thoughts to load the posts you’re interested in. If you would like to view any post in context simply click on the date in each post.

Some potential topics to explore: cop, ssm, 62ni, 64, ni, lithium, gamma, lead, shielding, mossbauer, (stable, troubles), (mw, plant), (positive, negative), hot, lenr, thermal, steam, electricity, hydrogen, lugano, parkhomov

  • Josh Guetzkow

    Great tool — thanks!

  • Matt Sevrens

    Currently they’re sorted by volume of usage. Would it be preferred to be alphabetically?

    • Brokeeper

      Is it possible for a check box for either/or?

  • Gerard McEk

    Very nice Matt. Extremely powerful, so much that my IPad can’t cope with it. It would be even nicer when also the questions to AR were included, but I am afraid that then also my PC can’t cope with it anymore. Thanks!

  • NT

    Very Impressive and useful, thanks for this database tool Matt…

  • Bob Greenyer

    Brilliant Work Matt, thanks for the heads up on this

    As I said, I applied for a UK NESTA grant on behalf of the MFMP at the beginning of 2014 to build an on-line mind map app using the leading big data analysis tool and data from the LENR-CANR site – No prizes for guessing why they felt it was not worth the support.

    https://goo.gl/5h3TXf

    Had we been able to do this, so many researchers would be further along – including us.

    I have said in the past – if the New Fire wasn’t invented by humans already – machine learning systems would have invented it in the near future.

    • Matt Sevrens

      This would be a fun next project to look into. I’ll start exploring and see what I can do. My main skills are in machine learning and data visualization, so maybe I could do something with word2vec (https://www.tensorflow.org/versions/r0.7/tutorials/word2vec/index.html) on this dataset and then visualize the results.

      • Bob Greenyer

        That would be cool.

        Matt, If you want to – I’ll try to arrange again to get funding to support this – maybe with the data mining tool I cite. Now that it is becoming clear how everything works – I am amazed at how much is in the LENR-CANR papers that should have been clues to those that paid attention.

        I try and get Jed Rothwell on-board.

        • Matt Sevrens

          Not sure I really have time immediately, but when I have time to experiment with it it’s not a matter of funding really. Just energy and focus. Also the fact that not all of the papers there seem to be available for download. Do you know if the ones not available for download are only accessible via payment?

          • Bob Greenyer

            I will ask

            • Matt Sevrens

              I can always start with whats available now to write the code and then when we get access to more papers we can just plug it in. One thing that might be interesting is to cluster multiple papers together based on similarity. Or I could create a flattened visual representation of topics using word2vec and t-sne that links to individual papers. There’s a lot of ways to approach it. I need to let the idea stew a bit.

              • Bob Greenyer

                Sounds like a plan.

  • Zephir

    Nice visualization, but I can see a similar mindset in LENR community, like in mainstream science: it deals with impressive, but useless details instead of real experiments. That is to say, the contemporary science is increasingly about itself, not about its subject of research.

  • http://www.animpossibleinvention.com/ Mats Lewan

    Brilliant!

  • Brokeeper

    Matt, awsome conceptual app. One slight suggestion: You may want to alphabetize the pre-selections. Thanks for the tool!

  • georgehants

    Wonderful and now we wait to see how much of this potential genius’s work is confirmed in the upcoming report.

  • Ged

    This is honestly impressive work.