Update README.md

README.md

@@ -1,23 +1,27 @@
 # X-MAP: Heterogeneous Recommendation at Large-Scale
 
-X-MAP is a large-scale heterogeneous recommender. X-MAP is built on top of Apache Spark and implemented in Python.
+X-MAP is a large-scale heterogeneous recommender which is built on top of Apache Spark and implemented in Python.
+
+## Features
+- Provides heterogeneous recommendation based on artificial **AlterEgos** of users across multiple application domains.
+- Any recommendation algorithm can be run on using these AlterEgos.
+- Provides formal privacy guarantees.
 
 
 ## Getting Started
 
 ### Prerequisites
+X-MAP requires `Python 3`, `Numpy 1.10.4`, `Apache Spark 1.6.1` pre-installed on your machine.
 
-X-MAP requires `Python 3`, `Numpy 1.10.4`, `Apache Spark 1.6.1` installed on your machine.
-
-#### Installing
+#### Installation
 Please refer to
-[Anaconda](https://www.continuum.io/downloads), [Apache Spark](http://spark.apache.org/) for installation.
+[Anaconda](https://www.continuum.io/downloads), [Apache Spark](http://spark.apache.org/) for installation instructions.
 
-#### Installing (Docker)
+#### Docker installation
 We also provide a docker image for your convenience.
 
-If you do not have `docker` and `docker-compose` installed on your machine, please check the [official configuration guidance](https://www.docker.com/products/overview) for installation.
-Otherwise, open your console and go to the `platform` folder, and execute the below-mentioned command to setup the corresponding docker image.
+If you do not have `docker` and `docker-compose` installed on your machine, please check the [official configuration guidance](https://www.docker.com/products/overview) for installation guidelines.
+Otherwise, open your console, go to the `platform` folder, and execute the below-mentioned command to setup the corresponding docker image.
 
 ```
 docker-compose build
@@ -34,16 +38,18 @@ You will find an egg file, located in `dist/xmap-0.1.0-py3.5.egg`, that you will
 
 ## Running the tests
 ### Prerequisites
-X-MAP uses [Amazon Dataset](https://snap.stanford.edu/data/web-Amazon.html) for its experiment. For current implementation, the input data follows the following format:
-`<userid>\t<itemid>\t<rating>\t<time>`.
+X-MAP is tested on real-traces of [Amazon Dataset](https://snap.stanford.edu/data/web-Amazon.html). For current implementation, the input data follows the following format:
+`<userid>\t<itemid>\t<rating>\t<timestamp>`.
+
+Note that the timestamp is required if you want to incorporate temporal behaviour which is also provided by the AlterEgos.
 
 ### Run X-MAP
 We provide here two demonstrations: `twodomain_demo.py` and `multidomain_demo.py`. You can also tune the parameters in the file `parameters.yaml`.
 
-Note that the scipt should run successfully under provided docker image. Please check your local parameters (e.g., directory path) when working on your own application.
+Note that the scipt should run successfully using the docker image that we provided. Please check your local system settings (e.g., directory path) while working with the application.
 
 #### Run X-MAP locally on a machine with 4 cores using docker
-A simple example of how to run X-MAP locally.
+A simple example of how to run X-MAP on a local machine.
 
 ```
 spark-submit --master local[4] \
@@ -86,4 +92,4 @@ rmse = evaluator.evaluate(predictions)
 print("Root-mean-square error = " + str(rmse))
 ```
 ## Support
-Please raise potential bugs on `github`. If you have an open-ended or a research related question, you can use: [X-MAP group] (https://groups.google.com/forum/#!forum/x-map)
+Please raise potential bugs on [github] (https://github.com/LPD-EPFL-ML/X-MAP/issues). If you have an open-ended or a research related question, you can post it on: [X-MAP group] (https://groups.google.com/forum/#!forum/x-map).