Title :
Reinforcement learning for information seeking
Author_Institution :
Sch. of Comput. & Math. Sci., Robert Gordon´´s Inst. of Technol., Aberdeen, UK
fDate :
6/21/1905 12:00:00 AM
Abstract :
Reinforcement learning may be used to provide a similar record of usage for digital media as “wear and tear” on books and so underpin heuristics that guide Web search. The domain-specific Cora search engine applies reinforcement learning to focus its spidering efforts. Reinforcement learning from past browsing habits extracted from Web logs may also help guide current visitors towards potentially interesting pages in a Web site
Keywords :
information resources; information retrieval; learning (artificial intelligence); search engines; Web logs; Web search; Web site; current visitors; digital media; domain-specific Cora search engine; heuristics; information seeking; interesting pages; past browsing habits; reinforcement learning; spidering efforts;
Conference_Titel :
Microengineering in Optics and Optoelectronics (Ref. No. 1999/187), IEE Colloquium on
Conference_Location :
Glasgow
DOI :
10.1049/ic:19990888