System and method for creating data to train a conversational bot
A system and method for creating input data to be used to train a conversational bot may include receiving a set of conversations, each conversation including sentences, classifying each sentence into a dialog act taken from a number of dialog acts, for each set of sentences classified into a dialog act, clustering the set of sentences into clusters based on the content (e.g. text) of the sentences, each cluster having a cluster name or label, and generating a language model based on the cluster labels. Slots may be identified in the sentences based in part on the dialog act classifications. A bot may be trained using data such as the slots, language model, and clusters.
1. A method for creating input di to be used to train a conversational bot, the method comprising:
receiving a set of conversations, each conversation comprising sentences;
classifying each sentence into a dialog act of a plurality of dialog acts, resulting in multiple sets of sentences, each set corresponding to a different dialog act;
for each set of sentences classified into a dialog act, clustering the set of sentences into clusters based on the content of the sentences, each cluster having a cluster label;
generating a language model based on the cluster labels, the language model taking as input a series of cluster labels and predicting a next cluster label; and
training a conversational bot using the language model.
2. The method of claim 1 , comprising identifying, in the sentences, slots based in part on the dialog act classifications, the slots used as input to a conversational bot creation process.
3. The method of claim 2 comprising identifying for a slot a set of possible slot values.
4. The method of claim 1 wherein the language model comprises a neural network.
5. The method of claim 2 , comprising training a conversational bot using the language model, clusters and slots.
6. A system for creating input data to be used to train a conversational bot, the system, comprising:
a memory; and
one or more processors configured to:
receive a set of conversations, each conversation comprising sentences;
classify each sentence into a dialog act of a plurality of dialog acts, resulting in multiple sets of sentences, each set corresponding to a different dialog act;
for each set of sentences classified into a dialog act, cluster the set of sentences into clusters based on the content of the sentences, each cluster having a cluster label;
generate a language model based on the cluster labels, the language model taking as input a series of cluster labels and predicting a next cluster label; and
train a conversational bot using the language model.
7. The system of claim 6 , wherein the one or more processors are configured to identify, in the sentences, slots based in part on the dialog act classifications, the slots used as input to a conversational bot creation process.
8. The system of claim 7 , wherein the one or more processors are configured to identify for a slot a set of possible slot values.
9. The system of claim 6 , wherein the language model comprises a neural network.
10. The system of claim 7 , wherein the one or more processors are configured to train a conversational bot using the language model, clusters and slots.
11. A method for generating data to be used to generate an automatic conversational bot, the method comprising:
for a set of transcripts comprising sentences, grouping the sentences into dialog acts resulting in multiple sets of sentences, each set of sentences corresponding to a different dialog act;
for each set of sentences grouped into a dialog act, grouping the set of sentences into clusters, each cluster having a cluster label;
training a language model based on the cluster labels, the language model taking as input a series of cluster labels and predicting a next cluster label; and
training a conversational bot using the language model.
12. The method of claim 11 , comprising identifying, in the sentences, slots based in part on the dialog act groupings, the slots used as input to a conversational bot creation process.
13. The method of claim 12 comprising identifying for a slot a set of possible slot values.
14. The method of claim 11 wherein the language model comprises a neural network.
15. The method of claim 12 , comprising training a conversational bot using the language model, clusters and slots.