Update documentation

docs/importers/Autotext.md

@@ -0,0 +1,25 @@
+---
+layout: default
+title: Auto Text
+nav_order: 3
+parent: Importers
+---
+# Auto Text Importer
+The Auto Text Importer uses VocabSieve's databases to extract appropriate sentences from your text automatically. Currently, it is not very configurable, and will only add sentences with only one unknown word based on available data. 
+
+Except for German and Luxembourgish, words starting with capital letters are considered proper names and automatically considered known to avoid them being added into Anki. 
+
+This is currently experimental and may not work as well as the other importers, since it has to make a lot more guesses. Please report any bugs you observe. 
+
+You may want to use this for two reasons:
+
+## Usage 1: Teaching VocabSieve about your progress
+Anki data is the primary source VocabSieve draws on when determining whether or not you know a word. By adding a lot of cards into Anki and marking the cards as known, you can tell VocabSieve about your progress and calibrate its database.
+
+0. Configure your tracking data sources. See [Tracking]({{site.baseurl}}/configuration/tracking.html), and wait for the known data to be ready. You can check if the data is ready by opening the Statistics window.
+1. Obtain articles, books, or transcripts of things you've read or watched or listened to. Be sure the file to be imported is either in an ebook format (.epub, .fb2, .mobi, etc) or are a text file (.txt)
+2. Use Import > Auto import from text, and select your file.
+3. Simple press "Lookup selected", and view the results. You can then decide whether to add the cards.
+
+## Usage 2: Asynchronous sentence mining
+If you don't have a lot of time to do active sentence mining, this may present an compelling alternative. However, for it to work well, VocabSieve needs to know quite well exactly which words you know. Therefore, it may require going through a lot of false positive (known target word) cards first anyways.

docs/index.md

@@ -10,7 +10,7 @@ Welcome to VocabSieve manual.
 ## Quickstart
 
 Go to [Installation]({{site.baseurl}}/installation.html) and [Configuration]({{site.baseurl}}/configuration.html) 
-if you are already familiar with immersion and Anki and just want to quickly set up VocabSieve.
+if you are already familiar with immersion and Anki and just want to quickly set up VocabSieve. If you already have VocabSieve set up, you may want to look at [Resources]({{site.baseurl}}/resources.html) to find dictionaries and other files to use with VocabSieve.
 
 ## Community support
 

docs/resources.md

@@ -1,7 +1,7 @@
 ---
 title: Resources
 layout: default
-nav_order: 6
+nav_order: 7
 ---
 
 # Resources for VocabSieve

docs/configuration/tracking.md → docs/tracking.md

@@ -1,8 +1,7 @@
 ---
 layout: default
 title: Vocabulary tracking
-nav_order: 1
-parent: Configuration
+nav_order: 6
 ---
 # Vocabulary tracking
 
@@ -12,7 +11,7 @@ By default, 100 points is needed for a non-cognate word to be classified as know
 
 | Event | Default score | Note |
 | ----- | ----- | ---- |
-| Seen  | 8 [1]| [Paper](https://core.ac.uk/download/pdf/323110125.pdf) assumes 12 exposures is needed to acquire a word |
+| Seen  | 8 | [Paper](https://core.ac.uk/download/pdf/323110125.pdf) assumes 12 exposures is needed to acquire a word |
 | Lookup | 15 | Max one lookup event per day | 
 | Mature Anki card target word | 80 | |
 | Mature Anki card context | 30 | |

docs/workflows.md

@@ -25,7 +25,7 @@ When you turn on the extension, you will notice that sentences are underlined in
 ### Built-in reader
 If you have an epub to read and would like to use VocabSieve in interactive mode while reading on your computer, you can use the built in reader, which can be accessed by the "Reader" button on the top bar.
 
-![](https://camo.githubusercontent.com/ea02ac411022c3f99b812cfb4c24ed9ba70ebea0c620fd68b56cb7df4e337768/68747470733a2f2f692e706f7374696d672e63632f766d3766727637702f6f75742e676966)
+![](https://i.postimg.cc/vm7frv7p/out.gif)
 
 ## Synchronous workflows (Videos)
 
@@ -38,8 +38,13 @@ You can use VocabSieve as a dictionary tool with [asbplayer](https://github.com/
 ## Asynchronous workflows (Ereaders)
 
 ### KOReader and Kindle
-If you have an ereader or phone with KOReader installed, you can use the [KOReader Importer]({{site.baseurl}}/importers/KOReader)
+If you have an ereader or phone with KOReader installed, you can use the [KOReader Importer]({{site.baseurl}}/importers/KOReader.html)
 
-Alternatively, if you use a Kindle, you can use the [Kindle Importer]({{site.baseurl}}/importers/Kindle). They have essentially the same features.
+Alternatively, if you use a Kindle, you can use the [Kindle Importer]({{site.baseurl}}/importers/Kindle.html). They have essentially the same features.
 
-![](https://camo.githubusercontent.com/3f08051e2956b5d20c14ffcb5fcc91c694c2fa61359315fa8ca9b5ea15f2e24e/68747470733a2f2f692e706f7374696d672e63632f35796a33566a50422f6f75742e676966)
+![](https://i.postimg.cc/5yj3VjPB/out.gif)
+
+## Automatic mining (Experimental)
+With the addition of the vocabulary tracking functionalities it is also now possible to automatically detect words you don't know from an article, book, or a podcast/video transcript. Transcripts can be generated using Whisper. To use this, you can select "Auto import from text" in the Import menu. See [Auto Text Importer]({{site.baseurl}}/importers/Autotext.html) for more information.
+
+At first, this is likely to result in a large number of false positives, i.e. words you know but VocabSieve thinks you don't. The best course of action here is to make sure your Anki query string in Config > Tracking matches either suspended or flagged cards to treat them as mature. This is a fast way to teach VocabSieve what words you really know.