Entity

Time filter

Source Type

Bangkok, Thailand

Boongoen T.,Royal Thai Air Force Academy | Iam-On N.,Mae Fah Luang University
Proceedings - International Carnahan Conference on Security Technology | Year: 2016

Resolving ambiguous and unknown identities is crucial to intelligence analysis in which fraud and deceptive names are frequently used by criminals and terrorists to make their activities unnoticeable. Typical approaches rely on the similarity measure of textual and other content-based characteristics, which are usually not applicable in the case of falsely-defined and unknown instances. This barrier can be overcome through link information presented in communication behaviors, financial interactions and social networks. Link-based similarity measures have proven effective for identifying similar problems in the Internet and publication domains. Inspired by this observation, the paper presents new link-based algorithms that do not only concentrate on link structure as adopted by the existing methods, but also bring link properties into consideration. Intuitively, links are weighted in accordance to their uniqueness. Their performance are experimentally evaluated with datasets related to terrorism and similar tasks, espcially the data collection extracted from evidence ontology that is used for investigation of unrest in southern Thailand. © 2015 IEEE. Source


Iam-On N.,Mae Fah Luang University | Boongoen T.,Royal Thai Air Force Academy | Garrett S.,Aispire Consulting Ltd. | Price C.,Aberystwyth University
IEEE Transactions on Pattern Analysis and Machine Intelligence | Year: 2011

Cluster ensembles have recently emerged as a powerful alternative to standard cluster analysis, aggregating several input data clusterings to generate a single output clustering, with improved robustness and stability. From the early work, these techniques held great promise; however, most of them generate the final solution based on incomplete information of a cluster ensemble. The underlying ensemble-information matrix reflects only cluster-data point relations, while those among clusters are generally overlooked. This paper presents a new link-based approach to improve the conventional matrix. It achieves this using the similarity between clusters that are estimated from a link network model of the ensemble. In particular, three new link-based algorithms are proposed for the underlying similarity assessment. The final clustering result is generated from the refined matrix using two different consensus functions of feature-based and graph-based partitioning. This approach is the first to address and explicitly employ the relationship between input partitions, which has not been emphasized by recent studies of matrix refinement. The effectiveness of the link-based approach is empirically demonstrated over 10 data sets (synthetic and real) and three benchmark evaluation measures. The results suggest the new approach is able to efficiently extract information embedded in the input clusterings, and regularly illustrate higher clustering quality in comparison to several state-of-the-art techniques. © 2011 IEEE. Source


Coo M.,Asian Institute of Technology | Pheeraphan T.,Royal Thai Air Force Academy
Construction and Building Materials | Year: 2015

In casting preplaced aggregate concrete (PAC), coarse aggregates are preplaced into formworks then grouts are injected to fill voids. This casting method depends on the ability of the grout to fill voids, which depends on grout workability and coarse aggregate shape and gradation. The effects of fly ash replacement and sand content on low W/B (0.33) PA grout properties is studied, along with the effects of coarse aggregate gradation on PAC mixtures. Significant factor effects and interactions are identified through statistical factorial design of experiment. Results show that inclusion of sand reduces fresh grout workability while fly ash replacement in binders compensates for the loss of workability. The increase in sand content increases mechanical strength of PAC, while coarse aggregate gradation has no significant effect in PAC mechanical strength. Optimized sand and fly ash proportions improve strength up to 43%, 49% and 90% of compressive, tensile and flexural strength respectively, compared to pure cement PAC mixtures. © 2015 Elsevier Ltd. All rights reserved. Source


Coo M.,Asian Institute of Technology | Pheeraphan T.,Royal Thai Air Force Academy
Construction and Building Materials | Year: 2016

A laboratory investigation on the effect of sand, fly ash, and limestone powder on mechanical properties of preplaced aggregate concrete (PAC) and shear strength of reinforced PAC beams without stirrups was conducted. Sand, fly ash and limestone powder is varied in each mixture to show their effects on the mechanical and structural strength properties of PAC. Ten reinforced PAC beams were cast, with PAC compressive strength ranging from 9.7 MPa to 35.8 MPa. From the same mix proportions, PAC slab members were cast to show an example application of PAC in the precast concrete industry. Beams were load tested in a four-point loading configuration with shear span to depth ratio (a/d) of 2.5, while slabs were loaded in equal 1/3 points. Load bearing performances of beams were evaluated based on load at first flexural cracking and ultimate shear load capacity. It was found that sand reduces workability of PAC grouts, but could be compensated by adding fly ash, while no significant fresh property effects were observed when cement was replaced with limestone powder. PAC mechanical strength had been shown to decrease as more cement was replaced with limestone powder. Modulus of rupture and ultimate shear strength of PAC beam members without stirrups can be designed using ACI 318-14 provisions. Finite element simulation using Vector2 can predict the ultimate shear strength of beam members. PAC example application through PAC slab members has shown similar performance at a lower concrete cost. © 2016 Elsevier Ltd. All rights reserved. Source


Iam-On N.,Mae Fah Luang University | Boongoen T.,Royal Thai Air Force Academy
Machine Learning | Year: 2013

Cluster ensembles or consensus clusterings have been shown to be better than any standard clustering algorithm at improving accuracy and robustness across various sets of data. This meta-learning formalism also helps users to overcome the dilemma of selecting an appropriate technique and the parameters for that technique. Since founded, different research areas have emerged with the common purpose of enhancing the effectiveness and applicability of cluster ensembles. These include the selection of ensemble members, the imputation of missing values, and the summarization of ensemble members. In particular, this paper is set to provide the review of different matrix refinement approaches that have been recently proposed in the literature for summarizing information of multiple clusterings. With various benchmark datasets and quality measures, the comparative study of these novel techniques is carried out to provide empirical findings from which a practical guideline can be drawn. © 2013, The Author(s). Source

Discover hidden collaborations