Topic: From Information Retrieval to Recommendation Systems
Speaker: Maria Mateva – FMI, Experian: bg.linkedin.com/pub/maria-mateva/16/a1/34b
Time: Wednesday, Feb 25, 19:00
Place: Faculty of Mathematics and Informatics, Hall 101
“Information retrieval” stands for the science behind search engines – how they are built and how they operate on large corpora of text documents to satisfy the information needs of humans. Maria will introduce this field by delving into the vector-space model and some aspects of tolerant retrieval.
“Recommender systems” are used to suggest items of interest to end users, e.g. friends to follow, products to buy online, music videos to watch, articles to read, etc. Decisions usually rest on the preferences history of the person. Our focus will be on content-based recommender systems, which serve to suggest items described with (or consisting of) text.
Maria provided some insights from her experience about the crossing point between the vector-space model and the content-based approach in recommendations. In the end, she presented latent semantic indexing – a solution to finding relations between the objects in large-dimensional data.
After the presentation we will head to a nearby bar for informal networking and beer.
Seating is limited and you need to register and take your FREE ticket here: