Keras-Distributed-Streaming

Distributed Keras model for making predictions of sentiment from Spanish sentences in stream context using Spark Streaming and Apache Kafka. The project has a script that allows you to enter phrases by command line and send them to a Kafka topic. Then Kafka sends the information to Spark Streaming where through the Jupyter Notebook shows the polarity of sentences on the screen.

Prerequisites

Files:

• Cardellino wor2vect file

Software versions:

• Spark 2.0.2 pre-build for Hadoop 2.7 or later
• Apache Kafka 2.11-0.10.2.0 binary file for Scala 2.11
• Python 3.5

Library versions:

• TensorFlow 1.2.1
• Keras 2.0.6
• Nltk 3.2.2
• Kafka-python

Execution

First you have to start a ZooKeeper server

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

Then you have to start a Kafka server

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

Next step is create a topic in Kafka named "test"

bin/kafka-topics.sh --create --zookeeper localhost:2181 --replication-factor 1 --partitions 1 --topic test

Later, you have to open Jupyter-notebook writing the next line

$SPARK_HOME/bin/pyspark --master local[*] --packages org.apache.spark:spark-streaming-kafka-0-8_2.11:2.0.2

Finally you only have to follow the instruction of the notebook.