# Default algorithms

A default algorithm is a function that, given a certain context, returns a sorted list of items.\
It can either be the result of applying an AI model or a very simple sort on datasets.

## Similar items <a href="#h_01jdktv2dprb8ecnvsgfey1vde" id="h_01jdktv2dprb8ecnvsgfey1vde"></a>

ABTasty uses a proprietary version of a TF-IDF (short for term frequency–inverse document frequency) algorithm to show the “seen with” items that really matters. NB.: *The “most-seen-together” items are considered to be the most similar items, but you can increase the precision of the model by filtering results by category or product type for example*

**Inputs:** similarTo:itemID

&#x20;

## Associated items <a href="#h_01jdktxpff4amzab05h1md1950" id="h_01jdktxpff4amzab05h1md1950"></a>

ABTasty uses a proprietary version of a TF-IDF algorithm to show the “purchased with” items that really matters. NB.: *The “most-bought-together” items are considered to be the most relevant complementary items, but you can increase the precision of the model by filtering results by brand, compatibility or cross-category filters for example*

**Inputs:** complementaryTo:itemID

&#x20;

## Associated category items <a href="#h_01jdkvng5w1jgaa4fmwkg9qn9y" id="h_01jdkvng5w1jgaa4fmwkg9qn9y"></a>

ABTasty uses a proprietary version of a TF-IDF algorithm to show items from the “purchased with” categories that really matters. NB.: *The “most-bought-together” items are considered to be the most relevant complementary items, but you can increase the precision of the model by filtering results by brand, compatibility or cross-category filters for example*

**Inputs:** complementaryTo:itemID

&#x20;

## Sorted items (top sells, views, recency, ratings,…) <a href="#h_01jdkvrg0rt7s29rsadzahve9d" id="h_01jdkvrg0rt7s29rsadzahve9d"></a>

ABTasty crosses your catalog with your pageviews and transactions data to enable you to create any weighted score and sort your items the way you want. NB.: *By default, all standard ranking algorithms (best sellers, best views, weighted scoresd…) are based on 30 days of data but this period can be easily adjusted*

**Inputs:** limit:integer

&#x20;

## Used items (Last seen, added, purchased) <a href="#h_01jdkvtb2sx2smhm7147v5h49k" id="h_01jdkvtb2sx2smhm7147v5h49k"></a>

ABTasty’s recos-tag stores in real-time in the local storage the items viewed, added and purchased NB.: *last viewed or purchased items can be directly collected and displayed, but this information is also used in order to trigger User affinity, or “similar items” or “complementary items” or “best sellers in selected categories” algorithms*

**Inputs:** sinceDays:integer

&#x20;

## User affinity items <a href="#h_01jdkvwfechcx4x7959khc7wxk" id="h_01jdkvwfechcx4x7959khc7wxk"></a>

ABTasty uses a proprietary scoring model returning for each user the most relevant items (based on his/her browsing, cart and purchases history).

&#x20;

## Forced items <a href="#h_01jdkvxe6jwagq71d5xhga7x55" id="h_01jdkvxe6jwagq71d5xhga7x55"></a>

ABTasty enables you to force at any rank in any recommendation a custom list of selected item(s)

&#x20;

## Tuned to emotion items <a href="#h_01jdkvyd9k3e5k7cevfvwhvrcy" id="h_01jdkvyd9k3e5k7cevfvwhvrcy"></a>

With EmotionsAI enabled, ABTasty detects users emotions in real-time to show the right items. NB.: *Work in Progress - don’t hesitate to ask for your position on the waitlist before it gets too long!*

**Inputs:** emotion:emotionID


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://docs.abtasty.com/recommendations-and-merchandising_deprecated/recos-and-merch-algorythms/default-algorithms.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.