Search

Utility is an important concept in Economics. A variety of applications consider utility in real-life situations, which has lead to the emergence of utility-oriented mining (also called utility mining) in the recent decade. Utility mining has attracted a great amount of attention, but most of the existing studies have been developed to deal with itemset-based data. Time-ordered sequence data is more commonly seen in real-world situations, which is different from itemset-based data. Since they are time-consuming and require large amount of memory usage, current utility mining algorithms still have limitations when dealing with sequence data. In addition, the mining efficiency of utility mining on sequence data still needs to be improved, especially for long sequences or when there is a low minimum utility threshold. In this paper, we propose an efficient Pro jection-based U tility M ining (ProUM) approach to discover high-utility sequential patterns from sequence data. The utility-array structure is designed to store the necessary information of the sequence-order and utility. ProUM can significantly improve the mining efficiency by utilizing the projection technique in generating utility-array, and it effectively reduces the memory consumption. Furthermore, a new upper bound named sequence extension utility is proposed and several pruning strategies are further applied to improve the efficiency of ProUM. By taking utility theory into account, the derived high-utility sequential patterns have more insightful and interesting information than other kinds of patterns. Experimental results showed that the proposed ProUM algorithm significantly outperformed the state-of-the-art algorithms in terms of execution time, memory usage, and scalability. [ABSTRACT FROM AUTHOR]

Recently, a new research field called utility-oriented mining has attracted great attention. However, previous studies have shown a limitation in that they rarely consider the inherent correlation of items among patterns. For example, considering the purchase behaviors of consumers, a high-utility group of products (w.r.t. multi-products) may contain several very high-utility products with some low-utility products. However, it is considered to be a valuable pattern even if this behavior/pattern may not be highly correlated, or even if it happens by chance. In light of these challenges, we propose an efficient utility-mining approach, called non-redundant Co rrelated high- U tility P attern M iner (CoUPM) by considering the positive correlation and profitable value. The derived patterns with high utility and strong positive correlation can lead to more insightful availability than those patterns that only have high profitable values. The utility-list structure is revised and applied to store the necessary information of both correlation and utility. Several pruning strategies are further developed to improve the efficiency for discovering the desired patterns. Experimental results showed that the non-redundant correlated high-utility patterns have more effectiveness than some other kinds of interesting patterns. Moreover, the efficiency of the proposed CoUPM algorithm significantly outperformed the state-of-the-art algorithm. [ABSTRACT FROM AUTHOR]

The natural language processing (NLP) denotes a technique used to process data such as text and speech. Some of the fundamental research in NLP includes the named entity recognition, which recognizes the named entities (i.e., persons and companies) from texts, the semantic parsing, which converts a natural language utterance to a logical form, and the co-reference resolution, which extracts the nouns (including pronouns and noun phrases) pointing to the same reference body. In this paper, we focus on the mention extraction and classification, proposing a neural-encoded mention-hypergraph model named the BILU-NEMH to extract the mention entities from a content. The proposed BILU-NEMH model combines a mention hypergraph model with the encoding schema and neural network. The proposed model can effectively capture the overlapping mention entities of an unbounded length. The proposed model was verified by the experiments, and the obtained experimental results showed that the proposed model achieved better performance and greater effectiveness than the existing related models on most standard datasets. [ABSTRACT FROM AUTHOR]

Periodic pattern mining is a popular data mining task, which consists of identifying patterns that periodically appear in data. Traditional periodic pattern mining algorithms are designed to find patterns in a single sequence. However, in several domains, it is desirable to discover patterns that are periodic in many sequences. An example of such application is market basket analysis. Given a database of sequences of transactions made by customers, discovering sets of items that are periodically bought by customers can help understand customer behavior. To discover periodic patterns common to multiple sequences, this paper extends the traditional problem of mining periodic patterns in a sequence. Two novel measures are defined called the standard deviation of periods and the sequence periodic ratio. Two algorithms are proposed to mine these patterns efficiently called MPFPS BFS and MPFPS DFS , which perform a breadth-first search and depth-first search, respectively. Because the sequence periodic ratio is neither monotone nor anti-monotone, these algorithms rely on a novel upper-bound called boundRa and two novel search space pruning properties to find periodic patterns efficiently. The algorithms have been evaluated on multiple datasets. Results show that they are efficient and can filter numerous non periodic itemsets to identify periodic patterns. [ABSTRACT FROM AUTHOR]

Graph neural network is a breakthrough in applying deep learning to non-Euclidean space. It is widely used for tasks such as social network analysis, molecular function inference, drug repositioning and protein modeling, achieving outstanding performance on relational models. Despite the great success of graph neural networks, most of them cannot be generalized to various scenarios. This is because graph information needs variation for different tasks and fixed models limit the flexibility of feature extraction. To address this challenge, we design a graph filter that can be adaptively adjusted according to graph tasks. This filter combines the multi-view strategy with the learnable quadratic frequency response function, using the crests of the quadratic functions to adaptively emphasize the required information. We further design the graph convolutional network model base on this adaptive filter, named AF-GCN. Extensive experiments are performed with 13 SOTA models on 12 different real-world datasets, including homogeneous and heterogeneous datasets in the node classification task, biological datasets, and social network datasets in the graph classification task. AF-GCN achieves state-of-the-art results in various scenarios. In addition, AF-GCN has high interpretability in the graph spatial domain. The development process from Graph Convolutional Networks (GCN) to AF-GCN has historical similarities with the development of Convolution Neural Networks (CNN). [ABSTRACT FROM AUTHOR]

This work aims to provide a novel hybrid architecture to suggest appropriate hashtags to a collection of orpheline tweets. The methodology starts with defining the collection of batches used in the convolutional neural network. This methodology is based on frequent pattern extraction methods. The hashtags of the tweets are then learned using the convolution neural network that was applied to the collection of batches of tweets. In addition, a pruning approach should ensure that the learning process proceeds properly by reducing the number of common patterns. Besides, the evolutionary algorithm is involved to extract the optimal parameters of the deep learning model used in the learning process. This is achieved by using a genetic algorithm that learns the hyper-parameters of the deep architecture. The effectiveness of our methodology has been demonstrated in a series of detailed experiments on a set of Twitter archives. From the results of the experiments, it is clear that the proposed method is superior to the baseline methods in terms of efficiency. [ABSTRACT FROM AUTHOR]

• An efficient EFUPM to discover the fuzzy high-utility patterns is proposed. • A Hadoop-based HFUPM is proposed to handle large-scale databases. • Two upper-bounds are then designed to early remove the unpromising candidates. • Experiments showed that the better performance can be obtained. Over the past decade, high-utility itemset mining (HUIM) has received widespread attention that can emphasize more critical information than was previously possible using frequent itemset mining (FIM). Unfortunately, HUIM is very similar to FIM since the methodology determines itemsets using a binary model based on a pre-defined minimum utility threshold. Additionally, most previous works only focused on single, small datasets in HUIM, which is not realistic to any real-world scenarios today containing big data environments. In this work, the fuzzy-set theory and a MapReduce framework are both utilized to design a novel high fuzzy utility pattern mining algorithm to resolve the above issues. Fuzzy-set theory is first involved and a new algorithm called efficient high fuzzy utility itemset mining (EFUPM) is designed to discover high fuzzy utility patterns from a single machine. Two upper-bounds are then estimated to allow early pruning of unpromising candidates in the search space. To handle the large-scale of big datasets, a Hadoop-based high fuzzy utility pattern mining (HFUPM) algorithm is then developed to discover high fuzzy utility patterns based on the Hadoop framework. Experimental results clearly show that the proposed algorithms perform strongly to mine the required high fuzzy utility patterns whether in a single machine or a large-scale environment compared to the current state-of-the-art approaches. [ABSTRACT FROM AUTHOR]

Document Information retrieval consists of finding the documents in a collection of documents that are the most relevant to a user query. Information retrieval techniques are widely-used by organizations to facilitate the search for information. However, applying traditional information retrieval techniques is time consuming for large document collections. Recently, cluster-based information retrieval approaches have been developed. Although these approaches are often much faster than traditional approaches for processing large document collections, the quality of the documents retrieved by cluster-based approaches is often less than that of traditional approaches. To address this drawback of cluster-based approaches, and improve the performance of information retrieval both in terms of runtime and quality of retrieved documents, this paper proposes a new cluster-based information retrieval approach named ICIR (Intelligent Cluster-based Information Retrieval). The proposed approach combines k-means clustering with frequent closed itemset mining to extract clusters of documents and find frequent terms in each cluster. Patterns discovered in each cluster are then used to select the most relevant document clusters to answer each user query. Four alternative heuristics are proposed to select the most relevant clusters, and two alternative heuristics for choosing documents in the selected clusters. Thus, eight versions of the proposed approach are obtained. To validate the proposed approach, extensive experiments have been carried out on well-known document collections. Results show that the designed approach outperforms traditional and cluster-based information retrieval approaches both in terms of execution time and quality of the returned documents. [ABSTRACT FROM AUTHOR]