Home

ecue

An Assessment of Case-Based Reasoning for Spam Filtering

Sun, 03/09/2008 - 12:18 — lorcan

Publication Type:

Conference Paper

Authors:

Sarah Jane Delany; Pádraig Cunningham; Lorcan Coyle

Source:

Proceedings of the Fifteenth Irish Conference on Artificial Intelligence and Cognitive Science (AICS'2004), Castlebar, Mayo, Ireland, p.9-18 (2004)

Keywords:

cbr; spam filtering; ecue

Abstract:

Because of the changing nature of spam, a spam filtering system that uses machine learning will need to be dynamic. This suggests that a case-based (memory-based) approach may work well. Case-Based Reasoning (CBR) is a lazy approach to machine learning where induction is delayed to run time. This means that the case base can be updated continuously and new training data is immediately available to the induction process. In this paper we present a detailed description of such a system called ECUE and evaluate design decisions concerning the case representation. We compare its performance with an alternative system that uses Naïve Bayes (NB). We find that there is little to choose between the two alternatives in cross-validation tests on data sets. However, ECUE does appear to have some advantages in tracking concept drift over time.

An Assessment of Case-Based Reasoning for Spam Filtering

Sun, 03/09/2008 - 12:18 — lorcan

Publication Type:

Journal Article

Authors:

Sarah Jane Delany; Pádraig Cunningham; Lorcan Coyle

Source:

Artificial Intelligence Review, Springer Science+Business Media B.V., Volume 24, Issue 3-4, Number 3--4, p.359-378 (2005)

Keywords:

cbr; spam filtering; ecue

Abstract:

Because of the changing nature of spam, a spam filtering system that uses machine learning will need to be dynamic. This suggests that a case-based (memory-based) approach may work well. Case-Based Reasoning (CBR) is a lazy approach to machine learning where induction is delayed to run time. This means that the case base can be updated continuously and new training data is immediately available to the induction process. In this paper we present a detailed description of such a system called ECUE and evaluate design decisions concerning the case representation. We compare its performance with an alternative system that uses Naïve Bayes. We find that there is little to choose between the two alternatives in cross-validation tests on data sets. However, ECUE does appear to have some advantages in tracking concept drift over time.

A Case-Based Technique for Tracking Concept Drift in Spam Filtering

Sun, 03/09/2008 - 12:18 — lorcan

Publication Type:

Conference Paper

Authors:

Sarah Jane Delany; Pádraig Cunningham; Alexey Tsymbal; Lorcan Coyle

Source:

The Twenty-fourth SGAI International Conference on Innovative Techniques and Applications of Artificial Intelligence (AI-2004), Springer, Queens' College, Cambridge, UK, p.3-16 (2004)

Keywords:

cbr; spam filtering; ecue

Abstract:

Clearly, machine learning techniques can play an important role in filtering spam email because ample training data is available to build a robust classifier. However, spam filtering is a particularly challenging task as the data distribution and concept being learned changes over time. This is a particularly awkward form of concept drift as the change is driven by spammers wishing to circumvent the spam filters. In this paper we show that lazy learning techniques are appropriate for such dynamically changing contexts. We present a case-based system for spam filtering called ECUE that can learn dynamically. We evaluate its performance as the case-base is updated with new cases. We also explore the benefit of periodically redoing the feature selection process to bring new features into play. Our evaluation shows that these two levels of model update are effective in tracking concept drift.

A Case-Based Technique for Tracking Concept Drift in Spam Filtering

Sun, 03/09/2008 - 12:18 — lorcan

Publication Type:

Journal Article

Authors:

Sarah Jane Delany; Pádraig Cunningham; Alexey Tsymbal; Lorcan Coyle

Source:

Knowledge-Based Systems, Elsevier, Volume 18, Issue 4-5, Number 4--5, p.187-195 (2005)

Keywords:

cbr; spam filtering; ecue

Abstract:

Spam filtering is a particularly challenging machine learning task as the data distribution and concept being learned changes over time. It exhibits a particularly awkward form of concept drift as the change is driven by spammers wishing to circumvent spam filters. In this paper we show that lazy learning techniques are appropriate for such dynamically changing contexts. We present a case-based system for spam filtering that can learn dynamically. We evaluate its performance as the case-base is updated with new cases. We also explore the benefit of periodically redoing the feature selection process to bring new features into play. Our evaluation shows that these two levels of model update are effective in tracking concept drift.

Syndicate content