jqwotos/hootmentor

Hi there 👋

Welcome to my profile!

• 💻 In the past, I was a software/developer engineer at Klipfolio, Hoppier, RBC Amplify, Unity, and RBC X where I primarily worked in full-stack development, while also venturing into Data Analytics, SRE, and Project Management.
• 🛠 I'm currently a Software Developer at CrowdStrike within the Charts and UX team.
• 🌱 Previously, I was learning how to build Real-time collaboration apps by creating a music collaboration site for my honors project supervised under Lou D. Nel

hootmentor's Issues

possible backend in go

Possibly in the future, we can convert the backend system to go lang for scalability and then just bridge back to python for tensorflow.

better training data

As discussed in #1 we should improve the training data so it's more suited to providing useful data back to the user.

1. Gather more relevant training data
2. Manually find some keywords such as ["income", "salary'], etc.. for different cases.
3. Using thesaurus scraper, gather more keywords based on manual keywords (and possibly just recursively do this to multiple levels)
4. Format the output sentences so the bot spits out tags in sentences.
5. Parse bot output tags into useable information.

This would allow the bot not just chat but also still be able to be useful in providing information and not just a small chatbot.

Keeping up with the TF.__newest__version Loop function needed

A future enhancement for adding new loop function replace line 99 to 121 as
shell Tensor is no longer iterable
IE:

            ##############################  Encoding Stage ####
            for i in range(0, self.n_encode_lstm_step):
                if i > 0:
                    tf.get_variable_scope().reuse_variables()

                with tf.variable_scope("LSTM1"):
                    output1, state1 = self.lstm1(wordvec_emb[:, i, :], state1)

                with tf.variable_scope("LSTM2"):
                    output2, state2 = self.lstm2(tf.concat([padding, output1], 1), state2)

            ############################# Decoding Stage ####
            for i in range(0, self.n_decode_lstm_step):
                with tf.device("/cpu:0"):
                    current_embed = tf.nn.embedding_lookup(self.Wemb, caption[:, i])

                tf.get_variable_scope().reuse_variables()

                with tf.variable_scope("LSTM1"):
                    output1, state1 = self.lstm1(padding, state1)

                with tf.variable_scope("LSTM2"):
                    output2, state2 = self.lstm2(tf.concat([current_embed, output1], 1), state2)

Reinforcement Learning goals #1

Fix the model into more suitable chatbot for recommendation