Atlas: Innovative Document Search
Atlas is novel document search tool that combines the power of AI with classic full text search. Documents are automatically analysed for topics, which then can be explored in an interactive virtual map.
Try it Now!
Simply test our new search tool for yourself and discover over 2 million Wikipedia articles from a new perspective:
New Features
Current search engines miss out on a lot of potential due to their traditional list-based result presentation:
Complete Overview
Millions of documents can be observed in a single overview. Every document corresponds to a point on a virtual map.Topic based Sorting
Documents with similar topic can be discovered in the immediate vicinity around another document.Interactive Filtering
Search results can be filtered by interactively zooming into a desired topic area.
How does it work?
Our document search system consists of two components: a traditional full-text search and an innovative topic analysis. Based on state of the art Deep Learning methods, we automatically derive topics for each document and position them on a virtual map.
1. Documents are analyzed for topics
Relevant keywords are extracted and evaluated for each and every document. Expensive training data is not required. The documents themselves are sufficient.
Apple II Series
The Apple II series is a family of home computers, one of the first highly successful mass-produced microcomputer products.
Table apple
Table apples or dessert apples are a group of apple cultivars grown for eating raw as opposed to cooking or cidermaking.
Pomacea diffusa
Pomacea diffusa, the spike-topped apple snail, is a species of freshwater snail, an aquatic gastropod mollusk in the family Ampullariidae, the apple snails.
2. Documents are grouped by topic on a virtual map
Documents with similar or connected keywords are arranged close to each other on a virtual map.
What are the Advantages?
Semantic Filtering via Zooming
When searching, only the documents that are currently visible on the map are listed as results. Since the documents are sorted by topic on the map, zooming into sections of the map will filter search results by topic.
Discover Related Documents
Similar documents are located next to each other on the map. When searching, you can discover important documents in the vicinity of a search result, even though they do not exactly match the query.
Exploration without any Search Queries
Discover any document by exploring the map. No search query required. Simply follow your imagination.
Specialized Workflows
Depending on the usage scenario complex workflows are conceivable, e.g., assigning documents of various topics through lasso selection, coloring documents or topics based on specific metrics or their processing status, and many more.
Example Scenarios
Discover the English Wikipedia from a new perspective. The following examples demonstrate how search results are arranged in different topics on a virtual map. Based on this map, you can both a quick overview over all topics and are able to inspect search results of individual topics.
Search regarding “Apple”
Discover various topics regarding the term “apple”, e.g.:
- Technology Products
like various Apple products, e.g., the Apple Watch and Apple S3 - Food and Groceries
like various apple related dishes, e.g., the apple strudel - Video Games
developed for Apple products
in comparison to a classic full-text search presented as a list (Wikipedia Search).
Search regarding “Concert”
Discover various topics regarding the term “concert”, e.g.:
- Events & Performances
like various artists and groups, e.g., Billy Joel - Halls and Venues
e.g., the Shanghai Concert Hall - Related Terminology
e.g., concert residency, house concert, virtual concert, and more.
in comparison to a classic full-text search presented as a list (Wikipedia Search).
We look Forward to Hearing from You
Send a Message
Call by Phone
+49-391-40594560
(Mon ‒ Fri, 9am ‒ 5pm)