Abstract
High-Utility Itemset Mining (HUIM) is a non-trivial task for analyzing databases such as, customer transactions, to reveal all itemsets that are of high importance. Over the past years, a number of traditional algorithms have been developed to mine High Utility Itemsets (HUIs). However, as a significant drawback, these algorithms often report a large number of HUIs, the majority of which lack any inherent correlations between the items in the HUIs. Such reported HUIs without inherent item correlation often are unreliable in decision making as they are usually of high utility by random chance of occurrence. To address this issue, this paper proposes the Correlated High Utility Itemset Miner (CHUIM). CHUIM employs the concept of productivity (in both normal and exclusive domains) to prune HUIs that occur by random chance without inherent item relationships. Experimental analysis on benchmark databases
show that, CHUIM is efficient and can prune the set of HUIs without inherent item correlations.
show that, CHUIM is efficient and can prune the set of HUIs without inherent item correlations.
| Original language | English |
|---|---|
| Title of host publication | 28th Pacific Asia Conference on Knowledge Discovery and Data Mining |
| Subtitle of host publication | Workshop on Utility-Driven Mining and Learning |
| Number of pages | 12 |
| Publication status | Published - 2024 |
| Event | 6th International Workshop on Utility-Driven Mining and Learning UDML 2024: at the 28th Pacific-Asia Conference on Knowledge Discovery and Data Mining PAKDD 2024 - Taipei International Convention Center (TICC), Taipei, Taiwan, Province of China Duration: 07 May 2024 → 07 May 2024 https://www.philippe-fournier-viger.com/utility_mining_workshop_2024/index.html https://onlinelibrary.wiley.com/page/journal/14680394/homepage/call-for-papers/si-2024-000154 (Call for papers for extended versions of workshop papers) |
Conference
| Conference | 6th International Workshop on Utility-Driven Mining and Learning UDML 2024 |
|---|---|
| Country/Territory | Taiwan, Province of China |
| City | Taipei |
| Period | 07/05/24 → 07/05/24 |
| Other | Utility-driven mining and learning from data has received emerging attentions from KDD communities due to its high potential in wide applications, covering finance, biomedicine, manufacturing, e-commerce, social media, etc. Some current research topics in utility-driven mining focused are for example to discover patterns of high value (eg, high profit) in large databases, analyzing/learning the important factors (eg, economic factors) in the data mining or machine learning process, and learning models that optimize some given utility functions. One of the popular applications of utility mining and learning is the analysis of large transactional databases to discover high-utility itemsets, which consist of sets of items that generate a high profit when purchased together. The workshop aims at bringing together academic and industrial researchers and practitioners from data mining, machine learning and other interdisciplinary communities, in the collaborative effort of identifying and discussing major technical challenges, recent results and potential topics on the emerging fields of Utility-Driven Mining and Learning. This workshop will focus on real world experiences, inherent challenges, as well as new research methods/applications. |
| Internet address |
|