Automatic XPath detectors are usually not a good choice when trying to figure out the Selenium locator for a specific web element. Especially expressions with numeric indexes (e.g. your li[3]) are likely to change if list/table items are removed, added or resorted.
The best way to locate an element is always by id, as this is always unique (unless you have invalid HTML). But your element doesn't have one, unfortunately.
For <a> elements, it's usually good to use the LinkText for locating the element, provided that a) it's unique and b) your site doesn't have a language toggle functionality, which usually changes all link texts.
Alternatively, you could use the tag name and class via CSS selector:
a.ng-binding
Still, it depends on the structure of your page whether this locator is unique or not. There are no general rules for finding the best locator, just good and bad strategies.