Just pushed beta version 128 for Chrome.
I've started incorporated a way to query using natural language for previously viewed papers through the extension. The advantage of this is obvious- if there is a paper that you want to quickly find you may be able to find it faster than using google. And you don't have to manually classify any papers you visit- this is done automatically just by previously viewing the papers. I need help testing it- challenging the interpreter with different query combinations allows me to train it to better to understand a wider range of query strategies.
To search, simply enter "LS" (without the quotes) followed by a space in the Chrome address bar. When you see the Lazy Scholar icon, any typing you do will automatically start searching after a short pause- no need to hit enter.
Here are some examples of what you can search for so far:
-Search "get last paper" or "last" and a reference to the most recent paper you viewed will display. Or, give it any number you'd like, such as "get last 10 papers".
-If you want to find the papers you visit the most, simply write something like "top papers", or "top" and it will give you an ordered list of results. Or, you can search within a topic, like "top papers about phosphorus", which only shows those papers you've visited most containing phosphorus in the title or as a keyword.
-You can also ask it to get papers from specific authors (currently just last name only), such as "last paper from Marks" to find the last paper you've visited by Marks.
I plan to expand what is searchable, but please let me know if you have ideas/requests.
Here is a video demo:https://www.youtube.com/watch?v=eG0vVTD8aSw