Dear All,
Artificial Intelligence usage in Libraries As Information retrieval tools in Universities i.e, Expert System, Natural Language Processing, Pattern Recognition and Robotics ? Being as Library users, Librarian, researcher, faculty member, expert uses your libraries please share your experience. If you have other aspects of Machine Learning/Deep learning please share
I will appreciated if you share any Master of PhD level study i.e (Thesis or Dissertation) on this topic or URL link/fulltext of published articles, Conference papers, Book Chapters, poster Presentation and etc.
Kind Regards,
-Yousuf
There are different aspects of machine learning, I think search algorithms like depth first search, breadth first search and A* ..etc. Also, Bayesian theory and analysis of big data can be aspect in machine learning.
Hi Ali, nowadays one can find a number of information retrieval and text mining applications in the field of library and information science. In particular, the applications of recommender systems can be identified as an emerging trend of this area. A number of initiatives have been taken place to implement journal recommender systems. In my view, almost all of these systems can be divided into two separate categories. One is content-based systems, in which the text similarities are compared between the input article (mostly abstract) and the corpora journal articles (trainingset). In addition to content filtering, some of these are associated with a component to just filter results based on author's publication needs (e.g. processing time, impact factor, indexing databases, etc.). However, none of these systems are able to determine the overall influence of these publication needs, yet merely filter the results generated by the content-based system by the second component. For example, Elsevier journal finder, Edanz journal selector, JANE, and Springer journal selector, they all fall into this category.
The other type of journal recommender systems combines content-based filtering with knowledge-based filtering. Here, the knowledge-based component determines the composite influence of author's publication needs using a similarity measure. Therefore, it could be more advanced than the first category. You can find an already published PhD dissertation which developed this kind of recommender system from the following link:
https://edoc.hu-berlin.de/handle/18452/20889
This dissertation is in English and free to download in its pdf version.
Journal recommender systems implemented so far work based on different similarity and supervised learning algorithms. For instance, BM25, kNN, VSM based models, and SVM are common. However, so far, I don't notice journal recommender systems based on deep learning techniques.
Apart from journal recommenders, one can also find studies on venue-based recommender sytems for conferences. Unlike in journal recommenders, here one can observe that they have also used collaborative-based filtering in addition to content-based filtering.
Intelligent knowledge representation, archiving and retrieval are some of these aspect where AI is being investigated.
- Badges
- Science topic