Deep distributed sentiment analysis on data stream (example)

Text stream sentiment analysis using a distributed deep learning approach based on Convolutional Neural Networks.

Language: Python 2.7

Streaming platform: Kafka

Distributed deep learning: Spark + BigDL

Authors

Mario Ruggieri

e-mail: mario.ruggieri@uniparthenope.it

Dependencies

• Intel BigDL with Spark: https://bigdl-project.github.io/master/#PythonUserGuide/install-from-pip/
• Kafka 0.8.2.2: https://kafka.apache.org/downloads

Usage

Starting Kafka ecosystem (bin is in kafka_2.11-0.8.2.2):

./bin/zookeeper-server-start.sh config/zookeeper.properties ./bin/kafka-server-start.sh config/server.properties

Starting producers:

python [HERE PATH TO kafka_producer.py] python [HERE PATH TO kafka_producer_for_word_prediction.py]

Stream testing with pretrained models:

${SPARK_HOME}/bin/spark-submit  \
--py-files ${PYTHON_API_ZIP_PATH},[HERE ABSOLUTE PATH TO cnn_stream_classifier.py]  \
--jars ${BigDL_JAR_PATH}  \
--conf spark.driver.extraClassPath=${BigDL_JAR_PATH} \
--conf spark.executor.extraClassPath=bigdl-0.2.0-SNAPSHOT-jar-with-dependencies.jar  \
--conf spark.executorEnv.PYTHONHASHSEED=${PYTHONHASHSEED} \
--packages org.apache.spark:spark-streaming-kafka-0-8_2.11:2.1.1 \
--driver-memory 10g  \
--executor-cores 4 \
--executor-memory 60g \
--num-executors 1 \
[HERE RELATIVE PATH TO cnn_stream_classifier.py] \
--action streaming_test \
--modelPath [HERE PATH TO model_for_sentiment]

${SPARK_HOME}/bin/spark-submit  \
--py-files ${PYTHON_API_ZIP_PATH},[HERE ABSOLUTE PATH TO lstm_word_prediction_single_out.py]  \
--jars ${BigDL_JAR_PATH}  \
--conf spark.driver.extraClassPath=${BigDL_JAR_PATH} \
--conf spark.executor.extraClassPath=bigdl-0.2.0-SNAPSHOT-jar-with-dependencies.jar  \
--conf spark.executorEnv.PYTHONHASHSEED=${PYTHONHASHSEED} \
--packages org.apache.spark:spark-streaming-kafka-0-8_2.11:2.1.1 \
--driver-memory 10g  \
--executor-cores 4 \
--executor-memory 60g \
--num-executors 1 \
[HERE RELATIVE PATH TO lstm_word_prediction_single_out.py]  \
--action streaming_test \
--modelPath [HERE PATH TO model_for_word_pred]

Training (generating models):

${SPARK_HOME}/bin/spark-submit  \
  --py-files ${PYTHON_API_ZIP_PATH},[HERE ABSOLUTE PATH TO cnn_stream_classifier.py] \
  --jars ${BigDL_JAR_PATH}  \
  --conf spark.driver.extraClassPath=${BigDL_JAR_PATH} \
  --conf spark.executor.extraClassPath=bigdl-0.2.0-SNAPSHOT-jar-with-dependencies.jar  \
  --conf spark.executorEnv.PYTHONHASHSEED=${PYTHONHASHSEED} \
  --driver-memory 10g  \
  --executor-cores 4 \
  --executor-memory 60g \
  --num-executors 1 \
  [HERE RELATIVE PATH TO cnn_stream_classifier.py] \
  --checkpoint_path [HERE PATH TO THE CHECKPOINT PATH] \
  --log_path [HERE PATH TO THE LOG PATH]

${SPARK_HOME}/bin/spark-submit  \
--py-files ${PYTHON_API_ZIP_PATH},[HERE ABSOLUTE PATH TO lstm_word_prediction_single_out.p] \
--jars ${BigDL_JAR_PATH}  \
--conf spark.driver.extraClassPath=${BigDL_JAR_PATH} \
--conf spark.executor.extraClassPath=bigdl-0.2.0-SNAPSHOT-jar-with-dependencies.jar  \
--conf spark.executorEnv.PYTHONHASHSEED=${PYTHONHASHSEED} \
--driver-memory 10g  \
--executor-cores 4 \
--executor-memory 60g \
--num-executors 1 \
[HERE RELATIVE PATH TO lstm_word_prediction_single_out.p] \
--checkpoint_path [HERE PATH TO THE CHECKPOINT PATH]
--log_path [HERE PATH TO THE LOG PATH]

License

Please read Apache 2.0 License file