Welcome, this blog has been built to keep you updated on latest news about Alternative activities. Regulary, blog posts will be published on specific topic concerning the software ecosystem and alternative possibilities in details.
We are happy to present to you this update. This blog post expose what is new in Alternative 1.0.0.
Today there's so much software solutions that it is easy to be out of your depth. Alternative's objectif is to list in a neutral way software options.
After the release of the beta version at the end of March 2020, user feedback has pushed us to continue in this way. This is where we start to work on user experience improvements.
One of the many feedbacks has been to display more information about the listed software. A challenging request due to the amount of scale changes involved. This is why we absolutely need to reduce the data extraction time from software websites.
A robot analyzes web pages and assists us in filling the database. It saves time when extracting them. Before, the extraction time lasted on average 3 minutes per field compared to 1 minute today. This let us extract more data to provide a better service. In the beta version, we have 12 possible data fields for each product. Now, more than 40 fields per product are possible to run Alternative. Irremediably, we observe a bouncing effect, but it is still a precious time saving.
Once data is extracted and checked, Alternative needs a persistant storage for them. We looked for a fast and efficient database with flexible possibilities to quickly upgrade the service. The software ecosystem is likely to be describe by a complex graph. A graph is a set of node linked between them. Nodes can represent softwares or functionnalities and relationship capabilities or links. That's why alternative is now running on neo4j noSQL graph oriented database.
Database graph - nodes and relationships
On top of the database, an Application Programming Interface (API) bridge the database with the website.
For more than 20 year Apache Lucene has been a reference in search engine industry. Alternative is using all the power of Lucene to provide better search results. Search is improved by the natural language processing of the query.
Search results are displayed ordered by similarity. This indicator is computed from the number of functionnality in common with the provided software or service.
A lot of time was spent delivering a better user experience on mobile. Now a side page is rendered when you click on site in the result list on mobile.
Alternative is compatible with your smartphones
From the side page you can find software details. As you can see information is now much more abundant :
This details are not always available. Sometime less information will be displayed.
Beside few improvements and bug correction, we will now focus on repopulate the database. One other major objective is to work on Alternative commercial part.