Difference between revisions of "Team:NTNU Trondheim/Software/iGEM Matchmaker"
| Line 31: | Line 31: | ||
<figure> | <figure> | ||
| − | <img class="img-responsive" style="max-width:500px" src="https://static.igem.org/mediawiki/2015/9/95/NTNU_Trondheim_Matchmaker.png" alt=" | + | <img class="img-responsive" style="max-width:500px" src="https://static.igem.org/mediawiki/2015/9/95/NTNU_Trondheim_Matchmaker.png" alt=""> |
| − | "> | + | |
<figcaption>The new version of the iGEM Matchmaker automatically matches teams based on the keyword analysis of their project abstract and website.</figcaption> | <figcaption>The new version of the iGEM Matchmaker automatically matches teams based on the keyword analysis of their project abstract and website.</figcaption> | ||
</figure> | </figure> | ||
| + | |||
<br> | <br> | ||
| Line 44: | Line 44: | ||
<p>Keywords are extracted from the project abstract and description pages using keyword extraction algorithms. Based on training data from the food and drug administration, <a href="https://github.com/zelandiya/maui">MAUI</a> uses machine learning to extract the most relevant keywords for the teams and measures the weight of the keyword.</p> | <p>Keywords are extracted from the project abstract and description pages using keyword extraction algorithms. Based on training data from the food and drug administration, <a href="https://github.com/zelandiya/maui">MAUI</a> uses machine learning to extract the most relevant keywords for the teams and measures the weight of the keyword.</p> | ||
| + | |||
| + | <figure> | ||
| + | <img class="img-responsive" style="max-width:500px" src="https://static.igem.org/mediawiki/2015/5/5f/NTNU_Trondheim_Matchmaker_visits.png" alt=""> | ||
| + | <figcaption>Number of unique team visits to the iGEM Matchmaker websites one week within the announcement of the 3.0 version.</figcaption> | ||
| + | </figure> | ||
<h3>Matchmaking</h3> | <h3>Matchmaking</h3> | ||
Revision as of 13:04, 18 September 2015
Software
iGEM Matchmaker
Software repository
Our software is available on the iGEM GitHub software repository!
We have released a new version of the iGEM Matchmaker. New features include automatic matchmaking of iGEM teams using machine learning and keyword extraction. The website has been upgraded to use Bootstrap 3.0.
Code
The code is available on GitHub.
Keyword analysis
Keywords are extracted from the project abstract and description pages using keyword extraction algorithms. Based on training data from the food and drug administration, MAUI uses machine learning to extract the most relevant keywords for the teams and measures the weight of the keyword.
Matchmaking
The principle of matchmaking is as follows:
- A team will be helped on a topic by a team for which this topic is more relevant.
- A team will be better at the main topics of their project, and will need help for the least important topics.