# imessage-chatbot

**Repository Path**: yohannlee/imessage-chatbot

## Basic Information

- **Project Name**: imessage-chatbot
- **Description**:  Recurrent neural network -- generates messages in your style of speech! Trained on imessage data. Sqlite3, TensorFlow, Flask, Twilio SMS, AWS.
- **Primary Language**: Unknown
- **License**: Not specified
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2020-03-26
- **Last Updated**: 2020-12-19

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# imessage-rnn

In-progress project: Building a neural network to learn an individual's style of speaking, and respond in their manner.

**Current State:**
Trained on my iMessages from the past year (many of which are with my mom). ~**60%** accuracy in letter-by-letter message generation. Disclaimer: I am not responsible for what the neural network generates!
[Live Chat](http://chat.anastassia.io/).

### Technologies

* Training messages obtained with **SQL**
* Recurrent Neural Network (RNN) built with **TensorFlow**
* **Flask** app responds with generated messages to a **Twilio SMS** webhook
* **AWS EC2**: Trained RNN on p2.xlarge instance, message generator deployed through t2.micro instance

### Implementation


#### 1. Extract Data

Data was extracted from the iMessage sqlite3 database, located in `/Library/Messages/chat.db`.

Getting all messages from yourself:
```sql
SELECT text FROM message WHERE is_from_me = 1;
```

Getting all messages from another person:
```sql
-- get their handle_id
SELECT * FROM handle WHERE id="(their_phone_number)";
-- use it to extract their messages
SELECT text FROM message WHERE handle_id = "(handle_id)" AND is_from_me = 0;
```

Save the ascii messages to a .txt file and format it (ie; deal with emojis).

#### 2. Process Data

With `dataset.py`, process the text data with parameters from `config.py` into _batches_. The input shape for the recurrent neural network is: (None, 32, 62, 256).

#### 3. Train the Recurrent Neural Network model

Now draw the rest of the owl

#### 4. Use the model to generate text

When a text is sent to our Twilio webhook, it hits out app, which generates a message and responds.

<p align="center">
<img src="/docs/message_ai.png" height="600"/>
<img src="/docs/webapp_ai.png" height="600"/>
</p>
<p align="center">Chatbot available through both SMS and online chat.</p>

Our app is very basic. It has two backend routes, one to respond to texts via Twilio and one that responds to messages on the [online app version](http://chat.anastassia.io/).

```python
@app.route("/sms", methods=['GET', 'POST'])
def sms_reply():
    resp = MessagingResponse()
    rnn_message = generate_message(459)
    resp.message(rnn_message)
    return str(resp)

@app.route("/message", methods=['GET'])
def message_reply():
    return str(generate_message(459))
```

The current step is to generate responses of varying natural lengths 🤔.

### Next Steps

- [x] Add a second layer to the model
- [x] Improve RNN performance with LSTM cells
- [x] In progress: Build web app version with **React**
- [x] In progress: Generate responses of varying lengths
- [ ] Look into Natural Language Processing (NLP)
- [ ] Increase training data