Ch2R: A Chinese Chatter Robot for Online Shopping Guide

Abstract

In this paper we present a conversational dialogue system, Ch2R (Chinese Chatter Robot) for online shopping guide, which allows users to inquire about information of mobile phone in Chinese. The purpose of this paper is to describe our development effort in terms of the underlying human language technologies (HLTs) as well as other system issues. We focus on a mixed-initiative conversation mechanism for interactive shopping guide combining initiative guiding and question understanding. We also present some evaluation on the system in mobile phone shopping guide domain. Evaluation results demonstrate the efficiency of our approach.

Mean length of utterance among Mandarin-speaking children with and without DLD

“…More longitudinal studies about MLU of Mandarin-speaking children should be conducted to explore the relationship between MLU and age. In terms of cross-sectional research, Huang and colleagues (2016) found significant differences between the 3-year-old and 4-year-old groups, and the MLU-w values of the typically developing 3- and 4-year-old children were 3.99 and 4.29, respectively. For children with a language delay, the MLU-w scores of the 3- and 4-year-old children were 3.33 and 3.85, respectively.…”

“…Previous studies have shown that MLU was a valuable measure for different languages and different age groups under different sampling procedures (Hewitt et al, 2005; Huang, et. al., 2016; Rice et al, 2010).…”

“…Some variables may affect the results of MLU values and its applicability such as age, maternal educational level, family socioeconomic status, number of utterances collected, and sampling procedure. Regarding age, children were aged 2 years and 6 months to 9 years in Rice and colleagues (2010), 6 years in Hewitt et al (2005), 4 to 7 years in Cheung (1998), and 3 and 4 years in Huang et al (2016). Thus, there has been a wide age range studied, and all of the research supports MLU as an applicable measure.…”

“…The results were difficult to interpret and compared to other studies. Huang et al (2016) developed a procedure for collecting, transcribing, coding, and analysing language samples and reported high transcribing and word segmenting agreement. They collected language samples from 142 Mandarin-speaking, typically developing children and Mandarin-speaking children with language delay (age range: 3–4) from Taipei City in Taiwan.…”

“…Regarding studies involving MLU5 in Chinese, Li (2014) indicated that the progression of MLU tended to stagnate after children reached 54 months of age, whereas performance on MLU5 continued to demonstrate significant progression for a further 6 months. Huang et al (2016) analysed the MLU5 of Mandarin-speaking, typically developing children and children with language delay aged 3 and 4 years. They found that typically developing children produced significantly higher MLU5 than children with language delay.…”