In recent years, many systems and approaches for recommending information,
products or other objects have been developed. In these systems,
often machine learning methods that need training input to acquire
a user interest profile are used. Such methods typically need positive
and negative evidence of the user�s interests. To obtain both kinds
of evidence, many systems make users rate relevant objects explicitly.
Others merely observe the user�s behavior, which fairly obviously
yields positive evidence; in order to be able to apply the standard
learning methods, these systems mostly use heuristics that attempt
to find also negative evidence in observed behavior. In this paper,
we present several approaches to learning interest profiles from
positive evidence only, as it is contained in observed user behavior.
Thus, both the problem of interrupting the user for ratings and the
problem of somewhat artificially determining negative evidence are
avoided. The learning approaches were developed and tested in the
context of the Web-based ELFI information system. It is in real use
by more than 1000 people. We give a brief sketch of ELFI and describe
the experiments we made based on ELFI usage logs to evaluate the
different proposed methods.
%0 Journal Article
%1 schwab00
%A Schwab, Ingo
%A Pohl, Wolfgang
%A Koychev, Ivan
%B IUI '00: Proceedings of the 5th international conference on Intelligent
user interfaces
%C New York, NY, USA
%D 2000
%I ACM
%K learning profiling recommender_systems
%P 241--247
%R http://doi.acm.org/10.1145/325737.325858
%T Learning to recommend from positive evidence
%X In recent years, many systems and approaches for recommending information,
products or other objects have been developed. In these systems,
often machine learning methods that need training input to acquire
a user interest profile are used. Such methods typically need positive
and negative evidence of the user�s interests. To obtain both kinds
of evidence, many systems make users rate relevant objects explicitly.
Others merely observe the user�s behavior, which fairly obviously
yields positive evidence; in order to be able to apply the standard
learning methods, these systems mostly use heuristics that attempt
to find also negative evidence in observed behavior. In this paper,
we present several approaches to learning interest profiles from
positive evidence only, as it is contained in observed user behavior.
Thus, both the problem of interrupting the user for ratings and the
problem of somewhat artificially determining negative evidence are
avoided. The learning approaches were developed and tested in the
context of the Web-based ELFI information system. It is in real use
by more than 1000 people. We give a brief sketch of ELFI and describe
the experiments we made based on ELFI usage logs to evaluate the
different proposed methods.
%@ 1-58113-134-8
@article{schwab00,
abstract = {In recent years, many systems and approaches for recommending information,
products or other objects have been developed. In these systems,
often machine learning methods that need training input to acquire
a user interest profile are used. Such methods typically need positive
and negative evidence of the user�s interests. To obtain both kinds
of evidence, many systems make users rate relevant objects explicitly.
Others merely observe the user�s behavior, which fairly obviously
yields positive evidence; in order to be able to apply the standard
learning methods, these systems mostly use heuristics that attempt
to find also negative evidence in observed behavior. In this paper,
we present several approaches to learning interest profiles from
positive evidence only, as it is contained in observed user behavior.
Thus, both the problem of interrupting the user for ratings and the
problem of somewhat artificially determining negative evidence are
avoided. The learning approaches were developed and tested in the
context of the Web-based ELFI information system. It is in real use
by more than 1000 people. We give a brief sketch of ELFI and describe
the experiments we made based on ELFI usage logs to evaluate the
different proposed methods.},
added-at = {2009-06-22T17:28:38.000+0200},
address = {New York, NY, USA},
author = {Schwab, Ingo and Pohl, Wolfgang and Koychev, Ivan},
biburl = {https://www.bibsonomy.org/bibtex/2cd5f88c7b17e6446f5c90965ed4c4d7c/lefteris8},
booktitle = {IUI '00: Proceedings of the 5th international conference on Intelligent
user interfaces},
doi = {http://doi.acm.org/10.1145/325737.325858},
interhash = {42e43956e90d446d706b010dfd6ad54c},
intrahash = {cd5f88c7b17e6446f5c90965ed4c4d7c},
isbn = {1-58113-134-8},
keywords = {learning profiling recommender_systems},
location = {New Orleans, Louisiana, United States},
pages = {241--247},
publisher = {ACM},
timestamp = {2009-06-22T17:28:40.000+0200},
title = {Learning to recommend from positive evidence},
year = 2000
}