I am pretty sre that by "NLP" you actually mean "Natural Language Processing"[0] rather than "Neuro Linguisting Programming"[1] (which is what your site says), but I must admit it would be pretty interesting if I was wrong.

[0] https://en.wikipedia.org/wiki/Natural_language_processing

[1] https://en.wikipedia.org/wiki/Neuro-linguistic_programming

Can't believe we made that mistake! Thank you so much for pointing that out.

Oh, believe it. It is a far more common mistake that you might think (probably because the term that isn't about programing artificial neural networks is the one that has the words neuro and programming in it).

The entire team (three of us) have been on the website multiple times a day and didn't notice at all. Only HN can find these things haha!

As a power user of Chrome, I've always wanted to go back to certain articles that I had read — but I was always faced with pages and pages of Chrome history.

So we built out DigitalBrain — a chrome extension that automatically will categorize what you're surfing and put it into different buckets+ based on what subject/topic it is using NLP. We also help you annotate and take notes within those articles.

Please let us know what you think, we're happy to answer questions about anything. Let's change how we use our browser history.

Here is the demo: https://youtu.be/hfbY13xXMd8

Sign-up for early access: https://digital-brain.co/

What specific nlp tools are you using? For example under what bucket ie scikitlearn, tensorflow, pytorch, etc and are they in house?

We're using a pretrained model on tensorflow. However, we're planning to make it in-house in the near future.

1. Can I store my data locally? 2. Are you selling information on my browsing? 3. Does the extension allow me to choose whether I infringe copyrights? e.g. can I still annotate/save/read SciHub papers? https://sci-hub.tw/

Thanks for your question!

1. So far we are saving all the personal data locally. So we don't have access to what you are browsing personally. In the future we are going to save the data to our servers for the sake of synchronizing in case you want to see your profile from another device. 2. We are not selling any personal information. 3. Our extension doesn't use any APIs for copyrights. The only thing our servers receive is a link that the server can scrape for necessary information.

+

It looks interesting. What are you using for category names? Can I add my own private categories?

We're using a specific NLP tool that helps with categorization; the topics/subtopics are actually quite deep.

We're currently building a feature that will help users build their own private categories.

maybe you should join forces with https://worldbrain.io/ :)

they just have full text search, but the rest of the product is great, it could really be a good synergy

Thank you so much for your recommendation. Memex looks great!

Seems super interesting. What tech are you guys using?

Thank you so much!

We're using a couple of NLP tools and the server is on Node.JS