Fix typos in documentation and minor improvements (asreview#632)

Co-authored-by: Bart-Jan Boverhof <[email protected]>
Co-authored-by: govertv <[email protected]>
Co-authored-by: Jonathan de Bruin <[email protected]>

README.md

@@ -9,7 +9,7 @@
 [![Build Status](https://img.shields.io/endpoint.svg?url=https%3A%2F%2Factions-badge.atrox.dev%2Fasreview%2Fasreview%2Fbadge%3Fref%3Dmaster&style=flat)](https://actions-badge.atrox.dev/asreview/asreview/goto?ref=master) [![Documentation Status](https://readthedocs.org/projects/asreview/badge/?version=latest)](https://asreview.readthedocs.io/en/latest/?badge=latest) [![DOI](https://zenodo.org/badge/DOI/10.5281/zenodo.3345592.svg)](https://doi.org/10.5281/zenodo.3345592) [![CII Best Practices](https://bestpractices.coreinfrastructure.org/projects/4755/badge)](https://bestpractices.coreinfrastructure.org/projects/4755)
 
 Systematic Reviews are “top of the bill” in research. The number of scientific
-studies is increasing exponentially in many scholarly fields. Performing a
+studies are increasing exponentially in many scholarly fields. Performing a
 sound systematic review is a time-consuming and sometimes boring task. The ASReview
 software is designed to accelerate the step of screening abstracts and titles
 with a minimum of papers to be read by a human with no or very few false
@@ -31,7 +31,7 @@ ASReview software implements two different modes:
 - **Simulate** :chart_with_upwards_trend: The simulation modus is used to measure
   the performance of the active learning software on the results of fully labeled systematic
   reviews. To use the simulation mode, knowledge on programming and bash/Command Prompt
-  is highly recommanded.
+  is highly recommended.
 
 ## Installation
 

docs/source/features/pre_screening.rst

@@ -76,7 +76,7 @@ Partly Labeled Data
 
 If you want to include decisions you've already made prior to setting up your
 project, you can upload a partly labeled dataset containg labels for part of
-the data and unlabeled recors you want to screen with ASReview. This might be
+the data and unlabeled records you want to screen with ASReview. This might be
 helpful if you switch from screening in another tool to screening with
 ASReview, or when updating an existing systematic review with more recent
 publications.
@@ -200,14 +200,13 @@ though relatively simplistic, it seems to work quite well on a wide range of
 datasets.
 
 The query strategy determines which document is shown after the model has
-computed the relevance scores. With certainty-based is selected the document
-with the highest relevance score is showed followed by the 2nd in line,
-etcetera, untill a new model is trained with new relevance scores. When
-uncertainty-based is selected, the most uncertain docuemtn is sampled
-according to the model (i.e. closest to 0.5 probability).  When random is
-selected, as it says, randomly select samples with no regard to model assigned
-probabilities. **Warning**: selecting this option means your review is not
-going to be accelerated by ASReview.
+computed the relevance scores. The three options are: certainty-based, mixed and 
+random. When certainty-based is selected, the documents are shown in the order of
+relevance score. The document most likely to be relevant is shown first. When 
+mixed is selected, the next document will be selected certainty-based 95% of the 
+time, and randomly chosen otherwise. When random is selected, documents are shown 
+in a random order (ignoring the model output completely). **Warning**: selecting 
+this option means your review is not going to be accelerated by using ASReview.
 
 The feature extraction technique determines the method how text is translated
 into a vector that can be used by the classifier. The default is TF-IDF (Term

docs/source/features/screening.rst

@@ -1,7 +1,7 @@
 Screening
 =========
 
-The screen in which you provide labels for records shown to you by the
+The user interface in which you provide labels for records shown to you by the
 software is kept as simple as possible. This is because ASReview wants you to
 focus on the content of the text so that you can make your decision as a true
 Oracle. You can access the following features during screening.
@@ -26,15 +26,15 @@ To make a decision:
 
 1. Open ASReview LAB.
 2. Open a project.
-3. Click on either the button Relevant or Irrelevant.
+3. Click on either the *Relevant* or *Irrelevant* button.
 4. The next record is presented. Repeat the process of labeling.
 
 .. figure:: ../../images/asreview_screening_asreview_label.png
    :alt: ASReview Screening
 
 .. warning::
 
-  If you are in doubt, think harder and take your time to make a decision, you
+  If you are in doubt, take your time to think on the decision, you
   are the oracle. Based on your input, a new model will be trained in the
   background. If you make decisions faster than the model needs for computing
   new relevance scores, you will simply be presented with the record next in
@@ -106,7 +106,7 @@ of your screening process so far. To open the statistics panel:
 In the top of the statistics panel the project name, authors and total number
 of records in the dataset are displayed.
 
-The pie chart on the presents an overview of how many relevant (green) and
+The pie chart presents an overview of how many relevant (green) and
 irrelevant (orange) records have been screened so far. Also, the total number
 of records screened is displayed, as well as the percentage screened relative
 to the total number of records in the dataset.
@@ -168,8 +168,8 @@ Via the hamburger menu in the left-upper corner you can:
 2. You can access the :doc:`Project Dashboard <post_screening>`.
 3. Navigate to the documention via the `HELP <https://asreview.readthedocs.io/en/latest/>`_ button.
 4. Provide feedback or `contribute <https://github.com/asreview/asreview/blob/master/CONTRIBUTING.md>`_ to the code.
-5. Donate to the ASReview project via the `ASReview crowdfunding platform <https://asreview.nl/donate>`_
-6. Quit the software (your progress was saved automatically)
+5. Donate to the ASReview project via the `ASReview crowdfunding platform <https://asreview.nl/donate>`_.
+6. Quit the software (your progress is saved automatically).
 
 
 .. _keybord-shortcuts:

docs/source/guides/activelearning.rst

@@ -37,7 +37,7 @@ to the user for the next iteration of the cycle;
 irrelevant. The newly labeled record is moved to the training data set and
 it’s back to the previous step.
 
-.. figure:: ../../images/RITL.png
+.. figure:: ../../images/RITL.jpg
    :alt: Researcher in the Loop
 
 The interaction with the human can be used to train a model with a minimum

docs/source/guides/sim_overview.rst

@@ -8,30 +8,30 @@ Why run a simulation?
 
 Doing simulations can be a great way to assess how well ASReview performs for
 your particular purposes. The user can run simulations on previously labeled
-datasets to see how much time the user could have saved using ASReview.
+datasets to see how much time is saved by using ASReview.
 
 Doing the simulation
 --------------------
 
-The ASReview simulation mode works through the dataset exactly like an
-ASReview user would, using the inclusions and exclusions in the dataset to
+The ASReview simulation mode iterates through the dataset exactly like an
+ASReview user would, using the inclusions and exclusions as included in the dataset to
 learn in the active learning cycle. In this way, the entire screening process
 is replicated.
 
 You can use the simulation mode that is provided with the ASReview package. It
-can be accessed directly from the command line, for example:
+can be accessed directly from the command line, for example like:
 
 .. code-block:: bash
 
 	asreview simulate MY_DATASET.csv --state_file myreview.h5
 
 This performs a simulation of a default active learning model, where
 ``MY_DATASET.csv`` is the path to the dataset you wish to simulate on and
-where ``myreview.h5`` is the file where the results will be stored.
+where ``myreview.h5`` is the file wherein the results will be stored.
 
 
-More detail on specific model and simulation settings can be found in the
-Simulation options section. For how to prepare your data, see
+More details on specific model and simulation settings can be found in the
+Simulation options section below. For how to prepare your data, see
 :doc:`../intro/datasets`.
 
 
@@ -61,7 +61,7 @@ performance of your review:
 	asreview plot myreview.h5
 	asreview plot DIR_WITH_MULTIPLE_SIMULATIONS
 
-For an example of results of a simulation study, see
+For an example of the results of a simulation study, see
 :doc:`simulation_study_results`.
 
 
@@ -86,7 +86,7 @@ specific set of starting papers, you can use ``--prior_idx`` to select the
 indices of the papers you want to start the simulation with.
 
 The ``--n_instances`` argument controls the number of records that have to be
-labeled before the model is retrained again and is set at 1 by default. If
+labeled before the model is retrained, and is set at 1 by default. If
 you want to reduce the number of training iterations, for example to limit the
 size of your state file and the time to simulate, you can increase
 ``--n_instances``.

docs/source/lab/exploration.rst

@@ -4,12 +4,12 @@ Exploration Mode
 The exploration mode can be used to explore to performance of the active
 learning software and the performance of :ref:`different algorithms
 <feature-extraction-table>` on already labeled data. In this mode relevant
-records are displayed in green and a recall curce can be obtained.
+records are displayed in green and a recall curve can be obtained.
 
 It is assumed you have already installed Python and ASReview. If this
 is not the case, see :doc:`../intro/installation`. Also, you should
 have created a :doc:`project<launch>` - the name is not
-relevant, but is adviced to have a explore-prefix.
+relevant, but is advised to have a explore-prefix.
 
 
 Upload a Benchmark Dataset
@@ -74,21 +74,10 @@ START reviewing
 Start reviewing the first 50, 100 or even 200 papers. Abstracts in green are
 relevenant papers and abstracts in black are irrelevant.
 
-For the *PTSD Trajectories* dataset you expect to find about 7 out of 38
-relevant papers after screening 50 papers, 19 after screening 100
-papers and 36 after 200 papers.
-
-For the *Virus Metagenomics* dataset you expect to find 20 out of 120 relevant
-papers after screening 50 papers, 40 after screening 100 papers
-and 70 after 200 papers
-
-For the *Software Fault Prediction* dataset you expect to find 25 out of 104 relevant
-papers after screening 50 papers, 48 after screening 100 papers
-and 88 after 200 papers.
-
-For the *ACEinhibitors* dataset you expect to find 16 out of 41 relevant
-papers after screening 50 papers, 27 after screening 100 papers and 32 after
-200 papers.
+- For the *PTSD Trajectories* dataset you expect to find about 7 out of 38 relevant papers after screening 50 papers, 19 after screening 100 papers and 36 after 200 papers.
+- For the *Virus Metagenomics* dataset you expect to find 20 out of 120 relevant papers after screening 50 papers, 40 after screening 100 papers and 70 after 200 papers
+- For the *Software Fault Prediction* dataset you expect to find 25 out of 104 relevant papers after screening 50 papers, 48 after screening 100 papers and 88 after 200 papers.
+- For the *ACEinhibitors* dataset you expect to find 16 out of 41 relevant papers after screening 50 papers, 27 after screening 100 papers and 32 after 200 papers.
 
 
 Upload Your own Data for Exploration

docs/source/lab/oracle.rst

@@ -92,7 +92,7 @@ While you review the documents, the software continuously improves its
 understanding of your decisions, constantly updating the underlying model.
 
 As you keep reviewing documents and providing more labels, the number of
-unlabeled docuemtns left in the dataset will decline. When to stop is left to
+unlabeled documents left in the dataset will decline. When to stop is left to
 the you. The `blogpost *ASReview Class 101* <https://asreview.nl/asreview-class-101/>`_
 provides some tips on stopping with screening.
 

docs/source/plugins/covid19.rst

@@ -12,17 +12,16 @@ guidelines to transparently support medical doctors. Medical guidelines rely
 on comprehensive systematic reviews. Such reviews entail several explicit and
 reproducible steps, including identifying all likely relevant papers in a
 standardized way, extracting data from eligible studies, and synthesizing the
-results into medical guidelines. They need to scan hundreds, or even thousands
-of COVID-19 related studies, by hand to find relevant papers to include in
-their overview. This is error prone and extremely time intensive; time we do
+results into medical guidelines. One might need to manually scan hundreds, or even thousands
+of COVID-19 related studies. This process is error prone and extremely time consuming; time we do
 not have right now!
 
 The software relies on :doc:`Active learning <../guides/activelearning>` which denotes the
 scenario in which the reviewer is labeling data that are presented by a
 machine learning model. The machine learns from the reviewers’ decisions and
 uses this knowledge in selecting the reference that will be presented to the
-reviewer next. In this way, the COVID-19 related papers are presented ordered
-from most to least relevant based on the input from the user. The goal of the
+reviewer next. In this way, the COVID-19 related papers are presented in an orderly manner,
+that is from most to least relevant based on the input from the user. The goal of the
 software is to help scholars and practitioners to get an overview of the most
 relevant papers for their work as efficiently as possible, while being
 transparent in the process.
@@ -56,7 +55,7 @@ In addition to the full dataset, there is a subset available of studies
 published after December 1st, 2019 to search for relevant papers published
 during the COVID-19 crisis.
 
-The datasets are updated in ASReview plugin shortly after the release by
+The datasets are updated in ASReview plugin shortly after a release by
 the Allen Institute for AI.
 
 Pre-print dataset
@@ -69,7 +68,7 @@ since January 1, 2020. The preprint dataset is updated weekly by the
 maintainers (Nicholas Fraser and Bianca Kramer) and then automatically updated
 in ASReview as well. As this dataset is not readily available to researchers
 through regular search engines (e.g. PubMed), its inclusion in ASReview
-provided added value to researchers interested in COVID-19 research,
+provides added value to researchers interested in COVID-19 research,
 especially if they want a quick way to screen preprints specifically.
 
 

docs/source/plugins/overview_plugins.rst

@@ -13,6 +13,6 @@ extensions of ASReview:
   papers, number of inclusions. `GitHub <https://github.com/asreview/asreview-statistics>`__
 - ``asreview-hyperopt``: Optimize the hyperparameters of the models in ASReview. `GitHub <https://github.com/asreview/asreview-hyperopt>`__
 
-If an extension is not on this list, or you make one and want it added to this
-list, make an issue on `github
+If an extension is not on this list, or you made one and you would like it to be added to this
+list, please initiate an issue on `Github
 <https://github.com/asreview/asreview/issues>`__.