Algolia's search propels 1,000s of retailers to Black Friday success
In the midst of the Black Friday shopping frenzy, Algolia soared to new heights, setting new records and delivering an ...
Chief Executive Officer and Board Member at Algolia
In the midst of the Black Friday shopping frenzy, Algolia soared to new heights, setting new records and delivering an ...
Chief Executive Officer and Board Member at Algolia
When was your last online shopping trip, and how did it go? For consumers, it’s becoming arguably tougher to ...
Senior Digital Marketing Manager, SEO
Have you put your blood, sweat, and tears into perfecting your online store, only to see your conversion rates stuck ...
Senior Digital Marketing Manager, SEO
“Hello, how can I help you today?” This has to be the most tired, but nevertheless tried-and-true ...
Search and Discovery writer
We are proud to announce that Algolia was named a leader in the IDC Marketscape in the Worldwide General-Purpose ...
VP Corporate Marketing
Twice a year, B2B Online brings together America’s leading manufacturers and distributors to uncover learnings and industry trends. This ...
Director, Sales Enablement & B2B Practice Leader
Generative AI and large language models (LLMs). These two cutting-edge AI technologies sound like totally different, incomparable things. One ...
Search and Discovery writer
ChatGPT, Bing, Bard, YouChat, DALL-E, Jasper…chances are good you’re leveraging some version of generative artificial intelligence on ...
Search and Discovery writer
Your users are spoiled. They’re used to Google’s refined and convenient search interface, so they have high expectations ...
Technical Writer
Imagine if, as your final exam for a computer science class, you had to create a real-world large language ...
Sr. SEO Web Digital Marketing Manager
What do you think of the OpenAI ChatGPT app and AI language models? There’s lots going on: GPT-3 ...
Search and Discovery writer
In the fast-paced and dynamic realm of digital merchandising, being reactive to customer trends has been the norm. In ...
Staff User Researcher
You’re at a dinner party when the conversation takes a computer-science-y turn. Have you tried ChatGPT? What ...
Sr. SEO Web Digital Marketing Manager
It’s the era of Big Data, and super-sized language models are the latest stars. When it comes to ...
Search and Discovery writer
Did you know that 86% of the global population uses a smartphone? The 7 billion devices connected to the Internet ...
Staff SME Business & Optimization - UI/UX
The Cloud Native Foundation is known for being the organization behind Kubernetes and many other Cloud Native tools. To foster ...
TL;DR Revamp your technical documentation search experience with DocSearch! Previously only available to open-source projects, we're excited ...
Senior Engineering Manager
Relevance – it’s what we’re all going for with our search implementations, but it’s so subjective that it ...
Technical Writer
What is faceted search? A critical Search UI feature, where users can click on facet values to filter their search results.
This is the second article in a three-part series of blog posts that describe the technical and data aspects of facets and faceted search. Here in part 2, we examine a variety of data structures applied to different facet use cases.
Like many front-end UX patterns, faceted search requires a precise and well-tailored facets JSON dataset. Luckily, there are many ways to represent facets in a JSON record. Most faceting is simple, only needing only a single-word filter to describe a product or service — an apple is a fruit, Apple is a brand. We can use a single attribute to represent “apple”:
{
"name": "apple",
"type": "fruit",
"seasonal_categories": ["fruits->late-summer", "fruits->autumn"]
}
or
{
"name": "iPhone 8 Gold Protection Screen",
"brand": "Apple",
"category": ["smartphone->protective-screens"]
}
The facets “type” and “brand” are single-dimension attributes. But facets can get more complicated: you can set them up as hierarchies or nested attributes. In this article, we’ll see how different facet data models achieve different front-end functionalities, all of which are central to faceted search and guided navigation:
We’ll go into four common ways to define facets and filters. Each one has its purposes and advantages:
The simplest facets are the most important to start with. They convey the essence of an object with a minimum of keywords. A shirt is either short-sleeved or long-sleeved. A movie has only one average rating. But simple facets can also have multiple values. A t-shirt can have one color or be a mix of red and blue. A movie can be a “romantic sci-fi comedy spoof”. Simple and exact tagging help bring back the best items to your users, giving them exactly what they were looking for. In our previous article on faceted search, we spoke about how facets are not only for filtering but also for searching.
Using JSON, simple facets are easy to represent:
{
"name": "bold t-shirt",
"desc": “Be bold, wear a t-shirt with only one color”,
"color": “white”,
"sleeves": "short"
},
{
"name": "Hippie Vest with Fringes",
"desc": “Be hip: get back to where you belong, peace + love”,
"color": ["red", “orange”, “yellow”, “green”, "blue"],
"sleeves": "long"
}
More advanced facets contain nested information. Nesting is all about structuring your data. Data structuring tells a story about an item. Let’s compare a single non-nesting example to different kinds of nesting.
Simple values:
{
"name": "Brad Pitt"
}
Simple nesting:
{
"first_name": "Brad",
"last_name": "Pitt"
}
Nesting with a bit more complexity:
{
"actors": [
{
"first_name": "Brad",
"last_name": "Pitt"
},
{
"first_name": "Scarlett",
"last_name": "Johanson"
}
]
}
* In this article, we use actors for searching and as facets to filter results by specific actors.
You don’t always need to structure, or nest, your data. With search, your main concern is finding the record, which is all about content, not structure. Finding “brad pitt” will work just as good with a simple or complex structure.
While structuring JSON may add human clarity to a given record, the engine can work with simple or complex structures. Structuring data always requires more thinking and more time to maintain, so it’s important to consider its value and purpose before overthinking it.
However, simple nesting enables you to anticipate search patterns. For example, if people rely on last names to find authors, it makes sense to create a “last_name” attribute.
{
"last_name": "King",
"first_name": "Stephen"
},
{
"last_name": "Shakespeare",
"first_name": "William"
}
More complex nesting can be used for display purposes, not just for filtering or search. One example is displaying itemized information about your items. Let’s compare two use cases, one that needs simple attributes, the other more complex nesting. A company selling DVDs does not necessarily need to overthink structure. They just need to make sure the record contains enough necessary information for its users to find a movie. Here’s a likely structure with only two attributes:
{
"title": "Avengers: Infinity War",
"cast": ["Robert Downey Jr", "Chris Hemsworth", "Mark Ruffalo", "Chris Evans", "Scarlett Johansson"]
}
On the other hand, a website that provides detailed information about a film, like IMDB or Netflix, needs to display more structured information:
{
"title": "Iron Man",
"cast": [
{
"first_name": "John",
"last_name": "Smith",
"birth_year": "1978",
"birth_place": "New York City, New York"
},
{
"first_name": "Robert",
"last_name": "Downey Jr",
"birth_year": "1965",
"birth_place": "New York City, New York"
}
]
}
This latter record contains data for an onscreen information box. The only caveat is that when your index contains complex nesting, you need to indicate to the engine which part of the structure to search. For example, you’ll need to designate “cast.first_name” and “cast.last_name” as searchable, and “cast.birth_year” as both searchable and filterable.
Many faceted search experiences offer users the ability to browse category and subcategory hierarchies, helping them to narrow or broaden their search. You can use hierarchies to create category-based menu systems. While designing and maintaining complex categories takes time, hierarchies create a great faceted search and navigational experience when done right. Take a look at how a supermarket can categorize its data to help people find the right online aisle:
[
{
"name": "lemon",
"categories.lvl0": "products",
"categories.lvl1": "products > fruits"
},
{
"name": "tomato",
"categories.level0": "products",
"categories.level1": "products > summer",
"categories.level2": "products > summer -> vegetables"
}
]
Booksellers also do it, with each author organized into different categories, some with multiple hierarchies:
{
"name": "Ursula K. Le Guin",
"categories": [
"level0": "Books",
"level1": ["Books > Science Fiction", "Books > Literature & Fiction"],
"level2": ["Books > Science Fiction > Time Travel", "Books > Literature & Fiction > Literary"]
]
}
Not all facets need to be managed by you and your company. Users have their own ideas about how to define and categorize your products and services. For example, you might classify “Breaking Bad” as a TV series and a Crime Drama. But users will have other ideas. Here, we group all user tags under the attribute “_tags”:
{
"title": "Breaking Bad",
"_tags": ["drugs", “father figure”, “cancer”, “midlife crisis”, “violent”, “not for kids” ]
}
And it’s not only users who add facet tags. Increasingly, machine learning algorithms and other forms of AI combine user behavior, product descriptions, market data, and more, to find helpful ways to tag items. For example, as users search, click, and view your catalog, AI auto-tagging can improve the categorization of your products by adding surprising alternatives. You can get age group, purpose, and other characteristics of products that you hadn’t thought about.
A popular example of this is image auto-tagging, where image detection algorithms generate a set of common tags for images. Websites like Pinterest that offer gifs and images make use of this kind of technology.
Imagine that users are consistently buying certain items before every major holiday. AI-generated tags might generate the following.
{
"name": "Scotch tape",
"_tags": ["Christmas", “Halloween”, “New Years”]
}
This information can be used to merchandise and manage your content, such as making sure that “Scotch tape” shows up high in the search results during holiday seasons.
Additionally, you can transform tags as facets into your records. Here, we’ve analysed the tag and created a new facet “holiday_accessory”:
{
"name": "Scotch tape",
"holiday_accessory": true
}
As described in our previous article of this series on faceted search, facets are useful for searching content. Having facets that match keywords entered in by users improves your search measurably. On a functional level, the engine doesn’t actually care that it is searching a category, a simple facet, or a non-facet. It looks where it is told to look. It’s therefore important to direct your engine to look at the specific attribute needed.
With this in mind, we discussed how complex nesting structure returns results that tell a story about an item in a record. Amazon returns results not only with product name, description, price, and an image. Its results also include ratings, shipping info, popularity, and categories.
The secret here is to put all that extra information into one searchable index. You can repurpose every attribute, using them for search, filtering, and/or display purposes. This avoids a second trip to another part of your database to retrieve additional information.
We’ll be speaking more about that in the next article of this series on faceted search.
Simple and complex facets create a faceted search and navigational experience based on browsing — with or without search. You can display facets in many ways — on the left-side of the screen, front and center, as drop downs, menus, and many other creative ways to make it easy for users to browse with single clicks.
But wherever they appear on the screen, and however you design them in your data, facets enable your users to click and browse and ultimately discover your full online catalog of products and services. Here, the technical point is that the underlying JSON structure must be well-tailored to support your front-end’s needs. This article hopes to have presented a solid foundation for common use cases in any sector or domain.
This is the second article in our Facets & Data series. Our focus in this series is technical, outlining the logic and facets data model of facet search.
Powered by Algolia Recommend