Air quality has impacts on our health and environment extremely. To monitor air pollutants, with the maturity of wireless sensor network, low-cost air sensors are deployed to trace pollution sources and detect personal exposure. Regular quality audit of deployed sensors is essential to ensure data quality of large-scale air monitoring networks. However, inspecting tremendous sensors by professional technicians regularly will take much human resources. This paper proposed the key sensor discovery for efficient and effective quality audit of large-scale air sensor network.

Recommender systems are common in e-commerce platforms in recent years. Recommender systems are able to help users find preferential items among a large amount of products so that users’ time is saved and sellers’ profits are increased. Cross-domain recommender systems aim to recommend items based on users’ different tastes across domains. While recommender systems usually suffer from the user cold-start problem that leads to unsatisfying recommendation performance, cross-domain recommendation can remedy such a problem. This article proposes a novel cross-domain recommendation model based on regression analysis, partial least squares regression (PLSR). The proposed recommendation models, PLSR-CrossRec and PLSR-Latent, are able to purely use source-domain ratings to predict the ratings for cold-start users who never rated items in the target domains. Experiments conducted on the Epinions dataset with ten various domains’ rating records demonstrate that PLSR-Latent can outperform several matrix factorization-based competing methods under a variety of cross-domain settings. The time efficiency of PLSR-Latent is also satisfactory.

Pictures can realize the impressions of texts for readers, especially for fairy tales. If we can present the fairy stories in the form of visual pictures, children will not only be more willing to concentrate their attention on but also easily perceive the underlying messages. This work aims to pre sent the fairy tales by transforming the texts to the visual form of pictures. To achieve such goal, we develop a system, VizStory, which consists of three steps. First, we investigate the narrative structure of the story to segment the whole story. Second, we select representative keywords for each segment. Third, through Web image search, we find suitable pictures to compose the visualization. Experimental results with a set of human study show VizStory can present the narrative of stories with 75% accuracy.

Live fish recognition in open aquatic habitats suffers from the high uncertainty in many of the data. To alleviate this problem without discarding those data, the system should learn a species hierarchy so that high-level labels can be assigned to ambiguous data. In this paper, a systematic hierarchical partial classification algorithm is therefore proposed for underwater fish species recognition. Partial classification is applied at each level of the species hierarchy so that the coarse-to-fine categorization stops once the decision confidence is low. By defining the exponential benefit function, we formulate the selection of decision threshold as an optimization problem. Also, attributes from important fish anatomical parts are focused to generate discriminative feature descriptors. Experiments show that the proposed method achieves an accuracy up to 94%, with partial decision rate less than 5%, on underwater fish images with high uncertainty and class imbalance.

Automatic video editing is receiving increasingly attention as the digital camera technology develops further and social media sites such as YouTube and Flickr become popular. Background music selection is one of the key elements to make the generated video attractive. In this work, we propose a framework for background music recommendation based on multi-modal latent semantic analysis between video and music. The videos and accompanied background music are collected from YouTube, and the videos with low musicality are filtered out by musicality detection algorithm. The co-occurrence relationships between audiovisual features are derived for multi-modal latent semantic analysis. Then, given a video, a ranked list of recommended music can be derived from the correlation model. In addition, we propose an algorithm for music beat and video shot alignment to calculate the alignability of recommended music and video. The final recommendation list is the combined result of both content correlation and alignability. Experiments show that the proposed method achieves a promising result.

One important function of current social networking services is allowing users to initialize different kinds of activity groups (e.g. study group, cocktail party, and group buying) and invite friends to attend in either manual or collaborative manners. However, such process of group formation is tedious, and could either include inappropriate group members or miss relevant ones. This work proposes to automatically compose the activity groups in a social network according to user-specified activity information. Given the activity host, a set of labels representing the activity's subjects, the desired group size, and a set of must-inclusive persons, we aim to find a set of individuals as the activity group, in which members are required to not only be familiar with the host but also have great communications with each other. We devise an approximation algorithm to greedily solve the group composing problem. Experiments on a real social network show the promising effectiveness of the proposed approach as well as the satisfactory human subjective study.

Score reduction is a process that arranges music for a target instrument by reducing original music. In this study we present a music arrangement framework that uses score reduction to automatically arrange music for a target instrument. The original music is first analyzed to determine the type of arrangement element of each section, then the phrases are identified and each is assigned a utility according to its type of arrangement element. For a set of utility-assigned phrases, we transform the music arrangement into an optimization problem and propose a phrase selection algorithm. The music is arranged by selecting appropriate phrases satisfying the playability constraints of a target instrument. Using the proposed framework, we implement a music arrangement system for the piano. An approach similar to Turing test is used to evaluate the quality of the music arranged by our system. The experiment results show that our system is able to create viable music for the piano.

In online social networking services, there are a range of scenarios in which users want to search a particular person given the targeted person one's name. The challenge of such people search is namesake, which means that there are many people possess the same names in the social network. In this paper, we propose to leverage the query contexts to tackle such problems. For example, given the information of one's graduation year and city, the last names of some individuals, one may wish to find classmates from his/her high school. We formulate such problem as the context-based people search. Given a social network in which each node is associated with a set of labels and given a query set of labels consisting of a targeted name label and other context labels, our goal is to return a ranking list of persons who possess the targeted name label and connects to other context labels with minimum communication costs through an effective subgraph in the social network. We consider the interactions among query labels to propose a grouping-based method to solve the context-based people search. Our method consists of three major parts. First, we model those nodes with query labels into a group graph which is able to reduce the search space to enhance the time efficiency. Second, we identify three different kinds of connectors which connecting different groups, and exploit connectors to find the corresponding detailed graph topology from the group graph. Third, we propose a Connector-Steiner Tree algorithm to retrieve a resulting ranked list of individuals who possess the targeted label. Experimental results on the DBLP bibliography data show that our grouping-based method can reach the good quality of returned persons as a greedy search algorithm at a considerable outperformance on the time efficiency.

Social networking services possess two features: (1) capturing the social relationships among people, represented by the social network, and (2) allowing users to express their preferences on different kinds of items (e.g. photo, celebrity, pages) through endorsing buttons, represented by a kind of endorsement bipartite graph. In this work, using such information, we propose a novel recommendation method, which leverages the viral marketing in the social network and the wisdom of crowds from endorsement network. Our recommendation consists of two parts. First, given some query terms describing user's preference, we find a set of targeted influencers who have the maximum activation probability on those nodes related to the query terms in the social network. Second, based on the derived targeted influencers as key experts, we recommend items via the endorsement network. We conduct the experiments on DBLP co-authorship social network with author-reference data as the endorsement network. The results show our method can achieve effective recommendations.

Mining repeating patterns from music data is one of the most interesting issues of multimedia data mining. However, less work are proposed for mining polyphonic repeating patterns. Hence, two efficient algorithms, A-PRPD (Apriori-based Polyphonic Repeating Pattern Discovery) and T-PRPD (Tree-based Polyphonic Repeating Pattern Discovery), are proposed to discover polyphonic repeating patterns from music data. Furthermore, a bit-string method is developed for improving the efficiency of the proposed algorithms. Experimental results show that the proposed algorithms, A-PRPD and T-PRPD, are both effective and efficient methods for mining polyphonic repeating patterns from synthetic music data and real data.

Piano reduction is a process that arranges music for the piano by reducing the original music into the most basic components. In this study we present an automatic arrangement system for piano reduction that arranges music algorithmically for the piano while considering various roles of the piano in music. We achieve this by first analyzing the original music in order to determine the type of arrangement element performed by an instrument. Then each phrase is identified and is associated with a weighted importance value. At last, a phrase selection algorithm is proposed to select phrases with maximum importance to arrangement under the constraint of piano playability. Our experiments demonstrate that the proposed system has the ability to create piano arrangement.

In this paper, we propose the emotion-based Impressionism slideshow system with automatic music accompaniment. While conventional image slideshow systems accompany images with music manually, our proposed approach explores the affective content of painting to automatically recommend music based on emotions. This is achieved by association discovery between painting features and emotions, and between emotions and music features respectively. To generate more harmonic Impressionism presentation, a linear arrangement method is proposed based on modified traveling salesman algorithm. Moreover, some animation effects and synchronization issues for affective content of Impressionism fine arts are considered. Experimental result shows our emotion-based accompaniment brings better browsing experience of aesthetics.

Spatial cognition concerns how human interpret spatial complexity. Cognitive maps are mostly used to test the spatial cognition. Analyzing cognitive maps drawn by students is helpful for teachers to understand students’ spatial cognitive ability and to draft geography teaching plans. Cognitive maps constitute the portfolios of spatial cognitive development. With the advance of e-learning technology, we can analyze portfolios of spatial cognitive development by spatial data mining of cognitive images. In this paper, we investigate the spatial co-orientation patterns for analyzing portfolios of spatial cognitive development. Spatial co-orientation patterns refer to objects that frequently occur with the same spatial orientation, e.g. left, right, below, etc., among images. We propose BFS-based approach for mining co-orientation patterns and utilize this as a tool for analyzing portfolios spatial cognitive development of students.

Mining multimedia data is one of the most important issues in data mining. In this paper, we propose an online one-pass algorithm to mine the set of frequent temporal patterns in online music query streams with a sliding window. An effective bit-sequence representation is used to reduce the processing time and memory needed to slide the windows. Experiments show that the proposed algorithm only needs a half of memory requirement of original music query data, and just scans the data once

In this paper, we propose an efficient algorithm, called MQS-change (changes of music query streams), to detect the changes of maximal melody structures in user-centered music query streams. Two music melody structures (set of chord-sets and string of chord-sets) are maintained and four melody structure changes (positive burst, negative burst, increasing change and decreasing change) are monitored in a new data structure MSC-list (a list of music structure changes). Experiments show that MQS-change algorithm is an online, single-pass approach to detect the changes of music melody structures over continuous music query streams

Computer music composition has been the dream of the computer music researcher. In this paper, we investigated the approach to discover the rules of music composition from given music objects, and automatically generate a new music object style similar to the given music objects. The proposed approach utilizes the data mining techniques to discover the rules of music composition characterized by the music properties, music structure, melody style and motif. A new music object is generated based on the discovered rules. To measure the effectiveness of proposed computer music composition approach, we adopted the method similar to the Turing test to test the discrimination between machine-generated and human-composed music. Experimental results showed that it is hard to discriminate. Another experiment showed that the style of generated music is similar to the given music objects.

Online mining of streaming data is one of the most important issues in data mining. In this paper, we proposed an efficient one-pass algorithm, called MFI-TransSW (mining frequent itemsets over a transaction-sensitive sliding window), to mine the set of all frequent itemsets in data streams with a transaction-sensitive sliding window. An effective bit-sequence representation of items is used in the proposed algorithm to reduce the time and memory needed to slide the windows. The experiments show that the proposed algorithm not only attain highly accurate mining results, but also run significant faster and consume less memory than existing algorithms for mining frequent itemsets over recent data streams.

Image mining is an important task to discover interesting and meaningful patterns from large image databases. We have previously introduced the spatial co-orientation patterns in image databases. Spatial co-orientation patterns refer to objects that frequently occur with the same spatial orientation, e.g. left, right, below, etc., among images. For example, an object P is frequently left to an object Q among images. We utilize the data structure, 2D string, to represent the spatial orientation of objects. In this paper, we propose an efficient algorithm, pattern-growth approach, for mining co-orientation patterns. An experimental evaluation with synthetic datasets shows the advantage and disadvantage between pattern-growth approach and the previous a priori-based approach.

In this paper, we proposed an online algorithm, called FQT-Stream (Frequent Query Trees of Streams), to mine the set of all frequent tree patterns over a continuous XML data stream. A new numbering method is proposed to represent the tree structure of a XML query tree. An effective sub-tree numeration approach is developed to extract the essential information from the XML data stream. The extracted information is stored in an effective summary data structure. Frequent query trees are mined from the current summary data structure by a depth-first-search manner.

The instructional activity model (IAM) is a general purpose model to generate an adaptive learning course which is compatible with the SCORM standard. IAM is composed of related activity tree (AT) nodes and capability nodes. Prerequisites are capabilities supposed to possess before learning an AT while contributions are capabilities after learning an AT. IAM model supports the adaptive learning sequencing by considering the relationships between AT and capability nodes. However, the IAM model does not take the transfer of learning into consideration. In this paper, we propose the mechanism to integrate the concept of learning transfer to the IAM model. In our proposed mechanism, the relationships between capabilities are considered based on the similarity measure between capabilities. The selection process of IAM model is also modified to reflect the relationships of capabilities.

With the growth of digital music, the development of music recommendation is helpful for users. The existing recommendation approaches are based on the users' preference on music. However, sometimes, recommending music according to the emotion is needed. In this paper, we propose a novel model for emotion-based music recommendation, which is based on the association discovery from film music. We investigated the music feature extraction and modified the affinity graph for association discovery between emotions and music features. Experimental result shows that the proposed approach achieves 85% accuracy in average.

Online, single-pass mining Web click streams poses some interesting computational issues, such as unbounded length of streaming data, possibly very fast arrival rate and just one scan over previously arrived click-sequencer In this paper, we propose a new, single-pass algorithm, called DSM-TKP (data stream mining for top-k path traversal patterns), for mining top-k path traversal patterns, where k is the desired number of path traversal patterns to be mined. An effective summary data structure called TKP-forest (top-k path forest) is used to maintain the essential information about the top-k path traversal patterns of the click-stream so far. Experimental studies show that DSM-TKP algorithm uses stable memory usage and makes only one pass over the streaming data.

We study the problem of mining frequent closed structures in a continuous, infinite-sized, and fast changing music melody stream. By modeling a music melody as a sequence of chord-sets, we propose an efficient algorithm FCS-stream (frequent closed structures of streaming melody sequences) for incremental mining of frequent closed structures in one scan of the continuous stream of chord-set sequences. An extended prefix-tree structure called TCS-tree (temporal closed structure tree) is developed for storing compact, essential information about the frequent closed structures of the stream. Results from our theoretical analysis and experimental studies with synthetic data show that the FCS-stream algorithm satisfies the main performance requirements, namely, single-pass, bounded memory, and real-time, for data stream mining.

With the growth of digital music, the personalized music filtering system is helpful for users. Melody style is one of the music features to represent user's music preference. We present a personalized content-based music filtering system to support music recommendation based on user's preference of melody style. We propose the multitype melody style classification approach to recommend the music objects. The system learns the user preference by mining the melody patterns from the music access behavior of the user. A two-way melody preference classifier is therefore constructed for each user. Music recommendation is made through this melody preference classifier. Performance evaluation shows that the filtering effect of the proposed approach meets user's preference.

Music style is one of the features that people used to classify music. Discovery of music style is helpful for the design of a content-based music retrieval system. In this paper we investigate the mining and classification of music style by melody from a collection of MIDI music. We extract the chord from the melody and investigate the representation of extracted features and corresponding mining techniques for music classification. Experimental results show that the classification accuracy is about 70% to 84% for 2-way classification.

The perception of spatial relationships among objects in a picture is one of the important selection criteria to discriminate and retrieve images in an image database system. The data structure called 2-D string, proposed by Chang et al., is adopted to represent the symbolic pictures. When there are a large number of images in the image database and each image contains many objects, the processing time for image retrievals is tremendous. It is essential to develop efficient access methods for these retrievals. In this paper, the efficient methods for retrieval by objects, retrieval by pairwise spatial relationships and retrieval by subpicture are proposed. All the methods are based on the superimposed coding technique.