IK
I. Kravcevs
info
Please Note
<p>This page displays the records of the person named above and is not linked to a unique person identifier. This record may need to be merged to a profile.</p>
2 records found
1
The project presents the design and implementation of a system for automatic parallelization of AWK programs. AWK remains a widely used language for text processing and data transformation. It is included as a standard utility tool on most Unix-like systems. The execution model of AWK is traditionally sequential, which limits scalability on multi-core hardware. The goal of this work is to investigate whether static program analysis can identify AWK scripts that can be executed in parallel and to integrate this capability into an AWK interpreter.
The proposed solution introduces a static analyzer that evaluates AWK programs based on variable dependencies, control flow, and other behaviors that impact data dependencies. The analyzer identifies reduction patterns for global variables and determines whether program semantics can be preserved under parallel execution. These results are then integrated into the interpreter, which enables deterministic multi-threaded execution.
The project adopts the MapReduce programming model to enable parallel execution of AWK. The main processing phase of a script is treated as the map stage, where independent partitions of the input are processed concurrently by multiple workers. Intermediate thread-local results are then combined in a reduce stage using aggregation strategies derived from static analysis. This model provides a structured way to preserve AWK’s sequential semantics in the parallelized environment.
The implementation was evaluated on a dataset of real-world AWK scripts and through performance benchmarks on large text-processing workloads. The results show that a significant subset of AWK programs can be parallelized automatically, achieving execution speedups and state-of-the-art AWK performance.
The project provides a practical path for improving efficiency in text-processing workflows. This work also demonstrates that scripting languages can often benefit from modern parallel execution techniques, extending their practical relevance and performance in data-processing tasks. ...
The proposed solution introduces a static analyzer that evaluates AWK programs based on variable dependencies, control flow, and other behaviors that impact data dependencies. The analyzer identifies reduction patterns for global variables and determines whether program semantics can be preserved under parallel execution. These results are then integrated into the interpreter, which enables deterministic multi-threaded execution.
The project adopts the MapReduce programming model to enable parallel execution of AWK. The main processing phase of a script is treated as the map stage, where independent partitions of the input are processed concurrently by multiple workers. Intermediate thread-local results are then combined in a reduce stage using aggregation strategies derived from static analysis. This model provides a structured way to preserve AWK’s sequential semantics in the parallelized environment.
The implementation was evaluated on a dataset of real-world AWK scripts and through performance benchmarks on large text-processing workloads. The results show that a significant subset of AWK programs can be parallelized automatically, achieving execution speedups and state-of-the-art AWK performance.
The project provides a practical path for improving efficiency in text-processing workflows. This work also demonstrates that scripting languages can often benefit from modern parallel execution techniques, extending their practical relevance and performance in data-processing tasks. ...
The project presents the design and implementation of a system for automatic parallelization of AWK programs. AWK remains a widely used language for text processing and data transformation. It is included as a standard utility tool on most Unix-like systems. The execution model of AWK is traditionally sequential, which limits scalability on multi-core hardware. The goal of this work is to investigate whether static program analysis can identify AWK scripts that can be executed in parallel and to integrate this capability into an AWK interpreter.
The proposed solution introduces a static analyzer that evaluates AWK programs based on variable dependencies, control flow, and other behaviors that impact data dependencies. The analyzer identifies reduction patterns for global variables and determines whether program semantics can be preserved under parallel execution. These results are then integrated into the interpreter, which enables deterministic multi-threaded execution.
The project adopts the MapReduce programming model to enable parallel execution of AWK. The main processing phase of a script is treated as the map stage, where independent partitions of the input are processed concurrently by multiple workers. Intermediate thread-local results are then combined in a reduce stage using aggregation strategies derived from static analysis. This model provides a structured way to preserve AWK’s sequential semantics in the parallelized environment.
The implementation was evaluated on a dataset of real-world AWK scripts and through performance benchmarks on large text-processing workloads. The results show that a significant subset of AWK programs can be parallelized automatically, achieving execution speedups and state-of-the-art AWK performance.
The project provides a practical path for improving efficiency in text-processing workflows. This work also demonstrates that scripting languages can often benefit from modern parallel execution techniques, extending their practical relevance and performance in data-processing tasks.
The proposed solution introduces a static analyzer that evaluates AWK programs based on variable dependencies, control flow, and other behaviors that impact data dependencies. The analyzer identifies reduction patterns for global variables and determines whether program semantics can be preserved under parallel execution. These results are then integrated into the interpreter, which enables deterministic multi-threaded execution.
The project adopts the MapReduce programming model to enable parallel execution of AWK. The main processing phase of a script is treated as the map stage, where independent partitions of the input are processed concurrently by multiple workers. Intermediate thread-local results are then combined in a reduce stage using aggregation strategies derived from static analysis. This model provides a structured way to preserve AWK’s sequential semantics in the parallelized environment.
The implementation was evaluated on a dataset of real-world AWK scripts and through performance benchmarks on large text-processing workloads. The results show that a significant subset of AWK programs can be parallelized automatically, achieving execution speedups and state-of-the-art AWK performance.
The project provides a practical path for improving efficiency in text-processing workflows. This work also demonstrates that scripting languages can often benefit from modern parallel execution techniques, extending their practical relevance and performance in data-processing tasks.
Efficient traffic forecasting is an important component of modern traffic management systems, enabling real-time route guidance and traffic control. Graph Neural Networks (GNN) have demonstrated state-of-the-art performance in this domain due to their ability to capture spatial and temporal dependencies in complex traffic data. However, GNNs typically require extensive historical data and are highly dependent on the specific road structure of the training region, posing challenges for their application in areas lacking such data. This study explores the transferability of GNN models in traffic forecasting, specifically how a GNN, trained in the region with long-horizon historical data, performs when applied to structurally different regional scenarios without historical data. The research investigates the impact of spatial differences between regions on the model's performance. The paper examines multiple metrics for regional similarity between training and transfer regions and shows their correlation with the transferred model's performance.
...
...
Efficient traffic forecasting is an important component of modern traffic management systems, enabling real-time route guidance and traffic control. Graph Neural Networks (GNN) have demonstrated state-of-the-art performance in this domain due to their ability to capture spatial and temporal dependencies in complex traffic data. However, GNNs typically require extensive historical data and are highly dependent on the specific road structure of the training region, posing challenges for their application in areas lacking such data. This study explores the transferability of GNN models in traffic forecasting, specifically how a GNN, trained in the region with long-horizon historical data, performs when applied to structurally different regional scenarios without historical data. The research investigates the impact of spatial differences between regions on the model's performance. The paper examines multiple metrics for regional similarity between training and transfer regions and shows their correlation with the transferred model's performance.