Hi, I’m Chow Yee Xiang and I will be presenting onthe functionalities of Amazon Textract and Amazon translate which areusing to extract text from photo or pictures and translate it into thelanguage that is requested.
So now I will show the demonstration of thesefunctions.
For example, this is the picture that we want to extract the textand translate into German language and the original language is English.
Before we start, take a look with the Python Script and the Python Script will requestfor some inputs including account ID which means email to identify the users, and the S3Bucket name which store documents, the documents name, origin language andthe targeted language.
Here are those languages that are supported in Amazon Translate.
Here is calling the boto3.
Boto3 is the AWS SDK for Python and it gets the Amazon Service.
Here is the Amazon Textract part where itget the function.
Here show the output of the extracted text and here shows the Amazon DynamoDB part, it puts the extracted text into dynamo DB forfuture work.
Here is the Amazon Translate part.
Now, we take a look with theresult of the Python Script.
As we can see, here are the inputs that we talked about before.
png” is the filename.
Here is the result of the extracted text.
Although the original documents is in table form, it is able to extract according to the column and row.
So hereis the translated result in German.
Lastly, we take a look with the DynamoDB, here is the extracted text that we extracted beforeand inserted into dynamo DB.
My name is Koay Yong Zhuang.
Now i am going to talk about Amazon Lex.
Amazon lex is a service for you to create chatbot.
First, create the bot but I’m going to show you the one I have createdhere.
Here you could create an intent which is actually an action.
In an intent, there will be the simple utterances where you could train your chatbot about theinput.
All these inputs will trigger this intend action.
Next, is the lambda initialization and validation.
Lambda is a service for youto run your code without thinking about servers.
This lambda function is going tobe called each time this intent is called.
You can include validation and dataprocessing in the lambda function.
Next, talking about the slots, here is whereyou prompt your questions for the data you want.
Here I asked for the location, check-in date, nights and the type of room.
Once all the questions have been answered by the user, the confirmation willbe prompted to get the confirmation from the user.
Once confirmed, it will go to thefulfillment whereby here I choose the same lambda function to process the datareceived above.
Here you can choose the response when the intent is ending.
After you have done the setting, you click save intent.
Then you build and publish this chatbot.
Next, speaking about channels, here are the list that you can integrate this chatbot to otherplatforms, such as Facebook, Kik, Slack and Twilio SMS.
For my case, I haveimplemented Facebook.
You need to add this callback URL to the Facebook APIfor Facebook to access this chatbot.
Next, I’m going to show you the lambdafunction.
Here you need to create a lambda function and I’m going to show you theone I have created here.
This is the code to handlethe lex intents which is written in Python 2.
I have uploaded this codeinto Google Drive, stated in the report for the needs.
In this code, I addedsave to DynamoDB after the intent is completed.
DynamoDB is the no SQLdatabase service provided by Amazon.
You need to change the AWS access key and secret access key to your own DynamoDB’s keys.
Remember to save your code.
Next, I’m going to show you the data save in theDynamoDB.
Here is it, all the data will be saved in this BookHotel table.
Thank you and that’s all for my part.
Hi Dr Cheng and my fellow classmates, I am Tan Sze Mei.
I will be presenting Amazon Sumerian integrated with Amazon Lex, which provide an Augmented Reality environment with a host-bot.
So, this is the dashboard of the Sumerian browser-based authoring tool.
The steps to be done for this scene is presented in the user manual.
So we will now test does it work.
Click publish and then copied the link.
Hello, my name is Christine.
I am a Sumerian host.
How can I help you? So when you want to speak, press the space bar and when you finish speaking, release the space bar.
Book a hotel.
What city will you be staying in? Chicago.
What day do you want to check in? Today.
How many nights will you be staying? Two.
What type of room would you like? Queen, King or Deluxe? Deluxe.
Okay, I have youdown for a two-night stay in Chicago, starting the 14th of April 2020.
Shall ibook the reservation? So this is the voice version.
So next, we will proceed to the text input version.
Hello, my name is Christine.
I am a Sumerian host.
How can I help you? [Please look into the white text box.
] what city will you be staying in? [Please look into the white text box.
] so basically the difference is you key in the textboxand the flow is the same as the voice version.
So, to save time I will stop here.
With that, I have finished the Sumerian AR demonstration in our project.
My name is Lian Yee Fu and now I will be presenting the blockchain service.
The blockchain service is used to track the reward point movement for data analysis on user loyalty and purchasing behaviour.
The demonstration will be done through the manual report.
Firstly, we have to create a new environment for AWS Cloud9 forconfiguring Blockchain.
Then, the network will be created with a small peer node.
Next, the client node is created to host Fabric CLI for administration purposes.
After that, channel is created on the client node.
Channel is a private “subnet” of communication between two or more specific network members, for the purpose of conducting private and confidential transactions.
Then the peer node is joined into the channel and the chaincode is installed and instantiated at the peer node to enable the interaction with the network’s shared ledger.
This is the example of query on dummy data from the shared ledger.
The first query retrieves reward point from user “a” and 90 points are returned.
The second query deduct 10 reward points from user “a” and transfer the points to merchant “b”.
The deduction is successful.
The third query again retrieves the reward point from user “a” and this time 80 points are returned because 10 points are deducted from the previous transaction.
Moreover, we can also monitor the point transaction and CPU, memory utilization of the network.
With that, I have finished the blockchain demonstration in our project.