rahulraj80 / gluonnlp-slot-filling

Demonstrate how to use GluonNLP to perform intent detection and slot filling

Joint Intent Classification and Slot Labeling by GluonNLP

Introduction

Intent classification and slot labeling are two essential problems in Natural Language Understanding (NLU). In intent classification, the agent needs to detect the intention that the speaker's utterance conveys. For example, when the speaker says "Book a flight from Long Beach to Seattle", the intention is to book a flight ticket. In slot labeling, the agent needs to extract the semantic entities that are related to the intent. In our previous example, "Long Beach" and "Seattle" are two semantic constituents related to the flight, i.e., the origin and the destination.

Essentially, intent classification can be viewed as a sequence classification problem and slot labeling can be viewed as a sequence tagging problem similar to Named-entity Recognition (NER). Due to their inner correlation, these two tasks are usually trained jointly with a multi-task objective function.

Here's one example of the ATIS dataset

Sentence	Tags	Intent Label
are	O	atis_flight
there	O
any	O
flights	O
from	O
long	B-fromloc.city_name
beach	I-fromloc.city_name
to	O
columbus	B-toloc.city_name
on	O
wednesday	B-depart_date.day_name
april	B-depart_date.month_name
sixty	B-depart_date.day_number

In this example, we demonstrate how to use GluonNLP to build a model to perform joint intent classification and slot labeling. We choose to finetune a pretrained BERT model. We use two datasets ATIS and SNIPS.

Requirements

mxnet
gluonnlp
seqeval

You may use pip or other tools to install these packages

Experiment

For the ATIS dataset, use the following command to run the experiment:

python demo.py --gpu 0 --dataset atis

It produces the final slot labeling F1 = 95.83% and intent classification accuracy = 98.66%

For the SNIPS dataset, use the following command to run the experiment:

python demo.py --gpu 0 --dataset snips

It produces the final slot labeling F1 = 95.76% and intent classification accuracy = 98.71%

Also, we train the models with three random seeds and report the mean/std

For ATIS

Models	Intent Acc (%)	Slot F1 (%)
Intent Gating & self-attention, EMNLP 2018	98.77	96.52
BLSTM-CRF + ELMo, AAAI 2019	97.42	95.62
Joint BERT, Arxiv 2019	97.5	96.1
Ours	98.66±0.00	95.88±0.04

For SNIPS

Models	Intent Acc (%)	Slot F1 (%)
BLSTM-CRF + ELMo, AAAI 2019	99.29	93.90
Joint BERT, Arxiv 2019	98.60	97.00
Ours	98.81±0.13	95.94±0.10