Value type: the bridge to a better DST model

Value type of the slots can provide lots of useful information for DST tasks. However, it has been ignored in most previous works. In this paper, we propose a new framework for DST task based on these value types. Firstly, we extract the type of token from each turn. Specifically, we divide the slots in the dataset into 9 categories according to the type of slot value, and then train a Ner model to extract the corresponding type-entity from each turn of conversation according to the token. Secondly, we improve the attention mode which is integrated into value type information between the slot and the conversation history to help each slot pay more attention to the turns that contain the same value type. Meanwhile, we introduce a sampling strategy to integrate these types into the attention formula, which decrease the error of Ner model. Finally, we conduct a comprehensive experiment on two multi-domain taskoriented conversation datasets, MultiWOZ 2.1 and MultiWOZ 2.4. The ablation experimental results show that our method is effective on both datasets, which verify the necessity of considering the type of slot value. ⇤ The first two authors contribute equally. Weiran Xu is the corresponding author. Sys: yes, the autumn house is on the east part of time, the prices are cheap and it is 4 stars. is there anything else you would like to know? Usr: no, i just want to book it for 2 people for 5 nights starting wednesday. Turn_label: hotel-area=east, hotel-book day=wednesday, hotel-people=2, hotel-book stay=5, hotel-princerange=cheap, hotel-stars=4 Sys: your friend has good taste. It is located at 02:00 rose crescent city centre, postcode cb23ll. Usr: i would like to book a table for 7 people on Monday at 15:15 please. Turn_label: restaurant-book day=monday, restaurant-book people=7, restaurant-book time=15:15 Sys: Usr: i would like a taxi from saint johns college to pizza hut fenditton. Turn_label: taxi-departure=saint johns college, taxi-destination=pizza hut fenditton