An intelligent automated assistant system engages with the user in an integrated, conversational manner using natural language dialog, and invokes external services when appropriate to obtain information or perform various actions. The system can be implemented using any of a number of different platforms, such as the web, email, smartphone, and the like, or any combination thereof. In one embodiment, the system is based on sets of interrelated domains and tasks, and employs additional functionally powered by external services with which the system can interact.

Intelligent Automated Assistant

A virtual assistant uses context information to supplement natural language or gestural input from a user. Context helps to clarify the user's intent and to reduce the number of candidate interpretations of the user's input, and reduces the need for the user to provide excessive clarification input. Context can include any available information that is usable by the assistant to supplement explicit user input to constrain an information-processing problem and/or to personalize results. Context can be used to constrain solutions during various phases of processing, including, for example, speech recognition, natural language processing, task flow processing, and dialog generation.

Using context information to facilitate processing of commands in a virtual assistant

Affective computing is currently one of the most active research topics, furthermore, having increasingly intensive attention. This strong interest is driven by a wide spectrum of promising applications in many areas such as virtual reality, smart surveillance, perceptual interface, etc. Affective computing concerns multidisciplinary knowledge background such as psychology, cognitive, physiology and computer sciences. The paper is emphasized on the several issues involved implicitly in the whole interactive feedback loop. Various methods for each issue are discussed in order to examine the state of the art. Finally, some research challenges and future directions are also discussed.

/pdf/affective-computing-a-review-52ggi0cvvb.pdf

Affective computing: a review

Automatic speech recognition for under-resourced languages: A survey

In 1998-1999, Prahalad and colleagues introduced the Base/Bottom of the Pyramid (BOP) concept in an article and a working paper. This article’s goal is to answer the question: What has become of the concept over the decade following its first systematic exposition in 1999? To answer this question, the authors conducted a systematic review of articles on the BOP, identifying 104 articles published in journals or proceedings over a ten-year period (2000-2009). This count excludes books, chapters, and teaching cases. The review shows that the BOP concept evolved dramatically following Prahalad’s original call to multinational enterprises (MNEs). De-emphasizing the role of MNEs over time, published BOP articles portray a more complex picture, with wide variations in terms of BOP contexts, of BOP initiatives, and of impacts of the BOP approach. A simple framework for organizing the reviewed articles helps discuss findings, identify the gaps that still exist in the literature, and suggest directions for future research.

https://papers.ssrn.com/sol3/Delivery.cfm/SSRN_ID2195659_code105013.pdf?abstractid=2193938&mirid=5&type=2

Reviewing a Decade of Research on the 'Base/Bottom of the Pyramid' (BOP) Concept

Virtually all work on topic modeling has assumed that the topics are to be learned over a text-based document corpus. However, there exist important applications where topic models must be learned over an audio corpus of spoken language. Unfortunately, speech-to-text programs can have very low accuracy. We therefore propose a novel topic model for spoken language that incorporates a statistical model of speech-to-text software behavior. Crucially, our model exploits the uncertainty numbers returned by the software. Our ideas apply to any domain in which it would be useful to build a topic model over data in which uncertainties are explicitly represented.

Topic Models over Spoken Language

One embodiment provides a method, including: receiving a request to merge at least one code branch with a first code branch, wherein the at least one code branch and the first code branch comprise different versions of software code; and simulating the merge via: identifying the differences between the at least one code branch and the first code branch; generating a differences list comprising a list of (i) added code snippets, (ii) deleted code snippets, and (iii) changed code snippets, wherein code snippets are clustered in the differences list by difference chunks comprising connected code snippets; and building a relationship graph comprising a plurality of nodes and a plurality of edges, wherein each of the plurality of nodes represent a difference chunk and wherein each of the plurality of edges comprise a relationship between two of the plurality of nodes; each of the plurality of edges comprise a cost vector.

Simulating and evaluating code branch merge

In this paper, we extend the concepts of audio-icons to provide auditory cues, for improving navigation in Spoken Web for low-literate population in rural India. The SWAicons are a suite of contextual auditory elements, which act as acoustic clues to a user. The SWAicons are expected to increase the user perception for the content and the structure of VoiceSite.

SWAicons: spoken web audio icons - design, implications and evaluation

Word clouds are extensively used to present a summary of the prominent words in a document on the World Wide Web. Such clouds give the user an idea about the content of the document. In this paper we present a mechanism to create and render an audio cloud for audio content. Such audio clouds are expected to provide a similar summary of the audio documents. They have wide applicability in various domains, especially for low-literate users who currently do not use the Internet but interact with audio-based systems.Detecting words from an audio content is challenging, especially if the audio is in languages for which a speech recognition system does not exist. We present a language-independent mechanism to detect frequently occurring words within an audio document. We then present four ways to render these words that form an audio cloud. The four prototypes for rendering the audio cloud are based on varying the amplitude, the voice quality, echo and the repetition of audio words. An evaluation study conducted across 32 users suggests that literate and low-literate users easily understand the concept of audio cloud.

Audio cloud: creation and rendering

We present a study of using embodied health information system for developing regions focusing on users not familiar with technology. We designed and developed a health information system with two gesture-based selection techniques: pointing to a screen and touching one's own body part. We evaluated the prototype in user study with 37 semi-literate and literate participants. Our results indicate a clear preference (76%) for touching in the healthcare domain. Based on our observations and user feedback, we present four design guidelines for developing embodied systems for the developing world: designing body-centric interaction to overcome literacy and technological proficiency barriers, addressing the misconceptions of system behaviors with users not familiar with technology, understanding effects of cultural constraints on interaction, and utilizing interactive virtual avatars to connect with the users.

Nitendra Rajput

Papers

Topic Models over Spoken Language

Simulating and evaluating code branch merge

SWAicons: spoken web audio icons - design, implications and evaluation

Audio cloud: creation and rendering

Body-touching: an embodied interaction technique for health information systems in developing regions