Bootstrapping Multilingual Intent Models via Machine Translation for Dialog Automation

Bootstrapping Multilingual Intent Models via Machine Translation for Dialog Automation Nicholas Ruiz, Srinivas Bangalore, John Chen Interactions, LLC Murray Hill, NJ, USA {nruiz,sbangalore,jchen}@interactions.com Abstract With the resurgence of chat-based dialog systems in consumer and enterprise appli- cations, there has been much success in de- veloping data-driven and rule-based natu- ral language models to understand human intent. Since these models require large amounts of data and in-domain knowledge, expanding an equivalent service into new markets is disrupted by language barriers that inhibit dialog automation. This paper presents a user study to evaluate the utility of out-of-the-box machine trans- lation technology to (1) rapidly bootstrap multilingual spoken dialog systems and (2) enable existing human analysts to under- stand foreign language utterances. We ad- ditionally evaluate the utility of machine translation in human assisted environ- ments, where a portion of the traffic is pro- cessed by analysts. In English→Spanish experiments, we observe a high potential for dialog automation, as well as the poten- tial for human analysts to process foreign language utterances with high accuracy. 1 Introduction With the present advances in natural language understanding and speech recognition technolo- gies, unprecedented opportunities have been cre- ated for realizing natural and sophisticated human- machine conversations to accomplish routine tasks. There has been a resurgence of speech/text- based conversation systems spanning multiple c⃝ 2018 The authors. This article is licensed under a Creative Commons 3.0 licence, no derivative works, attribution, CC- BY-ND. platforms, such as interactive voice recordings, chat, and SMS, owing to the availability of com- munication platforms that make it convenient to configure human-machine conversations. The po- tential opportunities of speech/text driven human- machine systems, or virtual agents, can only be re- alized if the user’s requests are understood by the virtual agent and acted upon appropriately. For practical applications, such conversational agents and speech/text analytics systems, the meaning of a sentence may be approximated as one or more actionable labels, or intents, associated with the input utterance. In such cases, the natu- ral language understanding (NLU) task is modeled as an intent classification problem. Although am- biguity is present in natural language, data-driven NLU systems have been successful in modeling user intents in many application domains. Many commercial and enterprise applications service customers from different geographic loca- tions and varying language proficiencies, requiring multilingual NLU for human-machine interaction. In order to deploy an intent classification system for a new language a new set of labeled training data is conventionally required. This data is often unavailable before a solution is deployed, instead requiring a human-driven dialog system depend- ing on intent analysts or live agents. In time, a sufficient amount of production data may be col- lected to build a data-driven intent model; however this approach is expensive to operate and ignores valuable knowledge present in other languages that could be used to build an initial model. In this paper, we evaluate the use of machine translation (MT) as a tool to bridge the knowledge present in one or more intent models for …

Content truncated to 500 words