The ever-increasing quantity of data generates enormous computational challenges in comparing and detecting matches in textual content. MIMOS Mi-AccelMorphe text analytics accelerator offers parallelised computatation with high-speed data processing on a heterogeneous platform.
MIMOS Mi-AccelMorphe is a text analytics accelerator tool designed for parallel data processing to compute record duplication and detect the similarity/dissimilarity of text/string data on an optimised heterogeneous platform. It enables high-speed text analytics operations by utilising a heterogeneous hardware acceleration platform that supports both Windows and Linux.
A text analytics accelerator tool for parallel data processing to compute record duplication and detect the similarity/dissimilarity of text/string.
Industries: Enterprise, Government
Mi-AccelMorphe addresses high volume data processing challenges
- Robust text mining services
- Accelerated processing algorithms
- Data scrambling
- Scalable heterogeneous framework
- Ultra-speed parallel data computation
- Transparent heterogeneous hardware support
Mi-AccelMorphe comprises the following features:
- Robust Text Mining Services
A service configurator framework provides optimum communication connectivity to data sources while allowing for scalability through an expandable parallel task scheduler.
- Accelerated Processing Algorithms
Mi-AccelMorphe rapidly processes and analyses incoming datasets by identifying matches on the same dataset and record linkages on different datasets such as complete, approximate, numeric and date matches and word similarity. It leverages on accelerated multi-core central processing unit (CPU) and graphics processing unit (GPU) algorithms.
- Data Scrambling
This feature ensures in-depth data privacy by performing data encryption and decryption of the processed datasets.
- Scalable Heterogeneous Framework
A scalable and configurable heterogeneous framework enables users to customise and extend functionalities via service plugin application programming interfaces (APIs).
The motivation of this project is to address the lack of reliable and efficient peat swamp forest data gathering and monitoring initiatives. A more systematic data gathering and monitoring system can assist in the immediate intervention, especially on the onset of fire triggered from long drought. The long term data gathered from the system will also give researchers a good understanding of the peat swamp ecosystem to enable them to devise a more systematic and sustainable peat swamp forest management in the future.
Currently, peat swamp forest monitoring is performed manually by the authorities or by reports from the villagers. A more reliable and efficient real-time remote monitoring can be achieved by way of an IoT-based system monitoring in the peat swamp forest areas. The sensory information from the sites is relayed back to a control centre through the clouds for data analytics and early warning purposes.