Instant answer

Abstract: Opioid as an addiction is a serious public health threat in the U.S., leads to massive deaths and other social problems. Medical treatment and mental supports are considering factors in rehabilitation process for opioid addicts. In this process families and friends play an important role in supporting and help the addict to stay clean. However, they may not know the best action to take due to lack of knowledge or certainty. Therefore, there are situations that addicts tend to use social media as a question/answering platform to seek answer for an inquiry. Unfortunately, It is often difficult to search over pages or different forums for a quick answer and it can be time-consuming, confusing and ultimately frustrating for the addicts. Hence, We propose a novel chatbot that is integrated with state-of-the-art deep learning techniques to retrieve an instant answer for a user’s query from Reddit social media. Our experiment illustrates that the chatbot provides answers in scenarios that there is no exact matched question in the discussion forums but there are questions with semantic similarities to the user query. Consequently, we illustrate real use cases where our chatbot retrieves responses from Reddit social media forums.

Abstract: A customer service automatic answering system and method aims to receive question contents entered by users, access the keywords of the question contents, search targeted topic contents in an answer database that match the keywords, and send answering data corresponding to the targeted topic contents to the network user for displaying, so that an automatic and instant answer may be performed and quick answer contents may be returned according to the question contents input by the network users.

Abstract: Software applications often use classification models to trigger specialized experiences for users. Search engines, for example, use query classifiers to trigger specialized "instant answer" experiences where information satisfying the user query is shown directly on the result page, and email applications use classification models to automatically move messages to a spam folder. When such applications have acceptable default (i.e., non-specialized) behavior, users are often more sensitive to failures in model precision than failures in model recall. In this paper, we consider model-selection algorithms for these precision-constrained scenarios. We develop adaptive model-selection algorithms to identify, using as few samples as possible, the best classifier from among a set of (precision) qualifying classifiers. We provide statistical correctness and sample complexity guarantees for our algorithms. We show with an empirical validation that our algorithms work well in practice.

Abstract: Architecture that automatically employs web search user query data to identify the places (e.g., locations, businesses) to which people are likely traveling, and then produces a heatmap visualization of the most popular places in the local area on a local map in a mapping application, which is then viewable in the vertical listing of the search results. This data can be utilized to rank local businesses in terms of popularity by knowing how many people are actually visiting the business as a function of date (and perhaps time). The web search data, which is used to understand the popular locations of a geographical area, includes signals such as searching for directions in the map application, and analyzing directions-related terms such as "From" and "To" in the search results. Another signal can be a location or business search which triggers an appropriate instant answer.