Mehrad docs update (#639)

docs/automations/extract-text-metadata.md

@@ -22,7 +22,7 @@ to get started with Automatic Metadata Extraction for Text.
     Identify and select the text fields from your dataset that you want to analyze.
     Also select the properties of the fields you wish to extract.
 
-    In the examble below we extract properties from the `best_answer` and `question` fields. For the `best_answer` field,
+    In the example below we extract properties from the `best_answer` and `question` fields. For the `best_answer` field,
     we display `word_count` and `topic_tag`, whereas for the `question` field we display `word_count`, `readability` and
     `question_type`.
 

docs/automations/index.md

@@ -6,7 +6,7 @@ hide:
 
 # :kolena-rocket-20: Advanced Usage
 
-This section contains tutorial documentation for Kolena automations.
+This section contains tutorial documentation for Kolena automation.
 
 <div class="grid cards" markdown>
 - [:kolena-properties-16: Automatically Extract Text Properties](./extract-text-metadata.md)

docs/automations/set-up-natural-language-search.md

@@ -12,13 +12,13 @@ or manually extracting and uploading corresponding search embeddings using a Kol
 ## Setting up Automated Embedding extraction
 
 ??? "Requirements"
-    - This feature is currenlty supported for Amazon S3 integrations.
+    - This feature is currently supported for Amazon S3 integrations.
     - Kolena requires access to the content of your images.
     Read [Connecting Cloud Storage: Amazon S3](../connecting-cloud-storage/) for more details.
     - Only account administrators are able to change this setting.
 
 Embedding extractions allow you to find datapoints using natural language or similarity between desired datapoints.
-To enable automated embedding, navigate to "Organization Settings" available on your profile menue, top right of the screen.
+To enable automated embedding, navigate to "Organization Settings" available on your profile menu, top right of the screen.
 Under the "Automations" tab, Enable the Automated Embeddings Extraction by Kolena option.
 
 <figure markdown>

docs/dataset/advanced-usage/dataset-formatting/computer-vision.md

@@ -122,12 +122,14 @@ bboxes = [
 Model results contian your model inferences as well as any custom metrics that you wish to monitor on Kolena.
 The data structure of model resutls is very similar to the structure of a dataset with minor differences.
 
-* Ensure your results are using the same unique ID feild (the `locator` for instance) you have selected for your dataset.
+* Ensure your results are using the same unique ID field (the `locator` for instance) you have selected for your dataset.
+
 * Use [`ScoredBoundingBox`](../../../reference/annotation.md#kolena.annotation.ScoredBoundingBox) or
 [`ScoredLabeledBoundingBox`](../../../reference/annotation.md#kolena.annotation.ScoredLabeledBoundingBox)
 to pass on your model inferences confidence score for each bounding box.
 * Use [`compute_object_detection_results`](../../../reference/experimental/index.md#kolena._experimental.object_detection.compute_object_detection_results)
-to compute your metrics that are supported by Kolena's [Object Detection Task Metrcis](../../advanced-usage/task-metrics.md#object-detection).
+to compute your metrics that are supported by Kolena's [Object Detection Task Metrics](../../advanced-usage/task-metrics.md#object-detection).
+
 * OR include the following columns in your results. The values for each of the columns is a [`List[ScoredLabeledBoundingBox]`](../../../reference/annotation.md#kolena.annotation.ScoredLabeledBoundingBox)
 
     | Column Name              | Description                                         |
@@ -182,7 +184,7 @@ The data structure of model results is very similar to the structure of a datase
 [`ScoredLabeledBoundingBox3D`](../../../reference/annotation.md#kolena.annotation.ScoredLabeledBoundingBox3D)
 to pass on your model inferences confidence score for each bounding box.
 * Use [`compute_object_detection_results`](../../../reference/experimental/index.md#kolena._experimental.object_detection.compute_object_detection_results)
-to compute your metrics that are supported by Kolena's [Object Detection Task Metrcis](../../advanced-usage/task-metrics.md#object-detection).
+to compute your metrics that are supported by Kolena's [Object Detection Task Metrics](../../advanced-usage/task-metrics.md#object-detection).
 
 !!! note
     Once you have constructed your `DataFrame` use the [`upload_object_detection_results`](../../../reference/experimental/index.md#kolena._experimental.object_detection.upload_object_detection_results)

docs/dataset/advanced-usage/index.md

@@ -33,4 +33,9 @@ This section contains tutorial documentation for advanced features.
     Automatically or manually extract embeddings from images to
     enable natural language and similar image search.
 
+- [:kolena-take-action-16: Programmatically Compare results](./quality-standard-results.md)
+
+    ---
+    Run model comparisons programmatically and add model improvements as requirements into your CI pipelines.
+
 </div>

docs/dataset/advanced-usage/quality-standard-results.md

@@ -3,7 +3,7 @@ icon: kolena/take-action-16
 
 ---
 
-# :kolena-take-action-20: Programatically Compare Models
+# :kolena-take-action-20: Programmatically Compare Models
 
 !!! example "Experimental Feature"
 
@@ -19,7 +19,7 @@ function, [`download_quality_standard_result`](../../reference/experimental/inde
 to download a dataset's quality standard result. This enables users to automate processes surrounding a Quality
 Standard's result.
 
-The return value is a multi-index DataFrame with indices `(stratificaiton, test_case)` and columns `(model, eval_config,
+The return value is a multi-index DataFrame with indices `(stratification, test_case)` and columns `(model, eval_config,
 metric_group, metric)`.
 
 <figure markdown>

docs/dataset/advanced-usage/task-metrics.md

@@ -70,7 +70,7 @@ supports [`average precision`](../../metrics/average-precision.md), [`precision`
 Kolena provides out-of-the-box aggregation options for your datapoint level evaluations that
 correspond with your desired metrics. For numeric evaluations you are able to
 select from `count`, `mean`, `median`, `min`, `max`, `stddev` and `sum` aggregations options.
-For categorical evaluations (class lable, boolean, etc) `rate` and `count` aggregation options are available.
+For categorical evaluations (class label, boolean, etc) `rate` and `count` aggregation options are available.
 
 The Kolena web application currently supports [`precision`](../../metrics/precision.md),
 [`recall`](../../metrics/recall.md), [`f1_score`](../../metrics/f1-score.md),

docs/dataset/core-concepts/index.md

@@ -55,8 +55,8 @@ You are able to select one or more fields as your ID field during the import pro
 Web App [:kolena-dataset-16: Datasets](https://app.kolena.com/redirect/datasets) or the
 SDK by using the [`upload_dataset`](../../reference/dataset/index.md#kolena.dataset.dataset.upload_dataset) function.
 
-**Meta data**: you can add additional informaiton about your
-datapoint simply by adding columns to the dataset with the metadaname and values in each row.
+**Meta data**: you can add additional information about your
+datapoint simply by adding columns to the dataset with the meta data name and values in each row.
 
 **Referenced Files**: each datapoint can contain a primary reference to a file stored on your cloud storage.
 Kolena automatically renders referenced files with column name `locator`. Other column names result in references to appear

docs/metrics/bertscore.md

@@ -95,7 +95,7 @@ BERT-precision and BERT-recall.
 
 In a more advanced implementation of BERTScore, extra steps are taken to finetune the metric. These include:
 
-1. Applying an "importance factor" to rare words so that the score weighs keywords moreso than words like "it", "as",
+1. Applying an "importance factor" to rare words so that the score weighs keywords more so than words like "it", "as",
 and "the".
 2. Rescaling the score such that it lies between 0 and 1 in practical use cases. Although the score already lies between
 0 and 1 in theory, it has been observed to lie between a more limited range in practice.

docs/metrics/diarization-error-rate.md

@@ -17,7 +17,7 @@ To simplest way to quantify the error in a candidate diarization is to measure t
 detections, and speaker confusions. These three elementary errors form the building blocks of diarization error rate.
 
 ??? example "False Alarm"
-    False alarm is duration of non-speech classified as speech — analagous to false positive diarizations. Using the
+    False alarm is duration of non-speech classified as speech — analogous to false positive diarization. Using the
     following hypothetical ground truth and inference diarization segments, let's calculate our false alarm duration.
 
     ```python
@@ -43,8 +43,8 @@ detections, and speaker confusions. These three elementary errors form the build
     doesn't exist in the ground truth diarization. Thus, our false alarm is equal to $3 + 1 = 4$ seconds.
 
 ??? example "Missed Detection"
-    Missed detection is the duration of speech classified as non-speech — analagous to a false negative in our
-    diarizations. Using the previous example, let's calculate our missed detection duration.
+    Missed detection is the duration of speech classified as non-speech — analogous to a false negative in our
+    diarization. Using the previous example, let's calculate our missed detection duration.
 
     ![Visualization](../assets/images/metrics-der-example1.png)
 
@@ -123,7 +123,7 @@ inference[Segment(23, 25)] = 'B'
 
     Our diarization error rate is 0.4.
 
-## Limitiations and Biases
+## Limitations and Biases
 
 Though DER provides a strong insight into the accuracy of speaker labels and predicted segments, it fails to
 pinpoint the specific components of a speaker diarization system that may cause it to perform poorly. As such,

docs/metrics/mean-squared-error.md

@@ -1,5 +1,5 @@
 ---
-description: How to calculate and interperet MSE (mean squared error) for regression ML tasks
+description: How to calculate and interpret MSE for regression ML tasks
 ---
 
 # Mean Squared Error (MSE)

docs/metrics/meteor.md

@@ -134,7 +134,7 @@ limitations.
 1. METEOR does not consider synonyms. Unlike embeddings-based metrics like [BERTScore](bertscore.md), it does not have
 a mechanism to quantify the similarity of words within the candidate and reference sentences. Thus, having two sentences
 like "She looked extremely happy at the surprise party." and "She appeared exceptionally joyful during the unexpected
-celebration." would yield a subobtimal score despite being very similar in meaning. That being said, METEOR has shown to
+celebration." would yield a suboptimal score despite being very similar in meaning. That being said, METEOR has shown to
 have a higher correlation with human judgement than both BLEU and ROUGE, making it *generally* better than the two.
 
 2. METEOR can fail on context. If we have two sentences "I am a big fan of Taylor Swift" (Reference) and "Fan of Taylor

docs/metrics/pr-curve.md

@@ -52,7 +52,7 @@ evaluating threshold; otherwise, it's positive.
 
 As the threshold increases, there are fewer false positives and more false negatives, most likely yielding high
 precision and low recall. Conversely, decreasing the threshold may improve recall at the cost of precision. Let's
-compute the precision and recall values at each threhold.
+compute the precision and recall values at each threshold.
 
 <center>