Showing papers by "Ali A. Minai published in 2018"

The Paths We Pick Together: A Behavioral Dynamics Algorithm for an HRI Pick-and-Place Task

Abstract: Behavioral dynamics models provide an observationally grounded basis for HRI algorithms and provide another tool for creating robust, natural, and interpretable HRI systems. Here, an HRI pick-and-place algorithm was implemented based on a behavioral dynamics model of human decision-making dynamics in an interpersonal pick-and-place task. Participants were able to complete the HRI pick-and-place task, we provide comparisons to HHI pick-and-place results.

Using Semantic Clustering And Autoencoders For Detecting Novelty In Corpora Of Short Texts

Abstract: Semantic analysis of text corpora is of broad utility, including for data from conversations, on-line chats, brainstorming sessions, comments on blogs, etc. – all of which are potentially interesting sources of information and ideas. In the present paper, we look at data from a large group brainstorming experiment that generated thousands of mostly brief statements. The ultimate goal is to detect which statements are semantically atypical within the overall corpus. In contexts such as spam detection or detection of on-line intrusions, autoencoders have been used successfully to separate typical from atypical data, and we consider this approach in the present paper. Texts are embedded in a semantic space obtained through topic analysis, and an autoencoder network is used to reconstruct each embedded text. The results show that, while difficulty of reconstruction is related to quantitative measures of atypicality in the embedding vector space, it is not well correlated with novelty assignments made by a human rater. However, this is not the case when the data is first clustered in the embedding space: The reconstruction error for each data cluster indicates that some clusters represent more novel data than others, and that the inverse size of the cluster and the mean reconstruction error of the texts in the cluster capture this well. In particular, autoencoders that enforce dimensionality reduction improve discrimination. The results also show that, in the reconstruction process, the autoencoder implicitly discovers the same clusters in the data that are discovered explicitly by an optimized k-means approach.

A Lexical Network Approach for Identifying Suicidal Ideation in Clinical Interview Transcripts

Abstract: Preventing suicide requires early identification of suicidal ideation. In this research, we propose an approach to evaluate whether an individual’s statements during a clinical interview can be classified as coming from a suicidal or non-suicidal mindset. To do so, we compare the statements with distinct lexical associative networks constructed from corpora of suicidal and control texts. Each node in these networks is a word, and the weight of the edge between every word pair indicates how strongly the words are associated in that corpus. Several metrics of association are evaluated in this work. Preliminary results show good classification performance with above 75% accuracy on novel test data.

Implicit Learning and Creativity in Human Networks: A Computational Model

Abstract: With rare exceptions, new ideas necessarily emerge in the minds of individuals through the recombination of existing ideas, but the epistemic repertoire for this recombination is supplied largely by ideas the individual has acquired from external sources, including interaction with peers. When agents hear new ideas and integrate them into their minds, they also implicitly create potential new ideas which can then become explicit as new ideas through later introspection. In this research, we use a multi-agent model to study such implicit learning in a social network and its relationship with the number of unique novel ideas actually expressed by agents in the network. We focus on the impact of two crucial factors: (1) The structure of the social network; and (2) The selectivity of agents in accepting ideas from their peers. We look at both latent ideas, i.e., those that are still implicit in the minds of individual agents, and novel expressed ideas, i.e., those that are expressed for the first time in the network. The results show that both network structure and the selectivity of influence have significant impact on the outcomes – especially in a system with misinformation.

Self-Organized Circle Formation around an Unknown Target by a Multi-Robot Swarm using a Local Communication Strategy

Abstract: Multi-robot systems have an innate advantage of enhancing system robustness in situations where the chances for noise in sensory data or faults in robotic agents are high. The distribution of the task at hand among multiple robots and coordination among these robotic agents can also make these systems more efficient. The formation problem for a multi-robot system is defined as the coordination of multiple robots to form and maintain a specific geometric pattern such as circle, line, or square. This paper focuses on circular pattern formation through self-organization in multi-robot systems. A detailed algorithm for circle formation around an unknown target using exceptionally simple robots is presented along with the results. In this algorithm, we adopt a communication-based multi-agent model where each robotic agent interacts through minimal, though explicit, communication within very limited range and with limited capabilities. This algorithm is entirely distributed and does not require any manual intervention or centralized information. Extensive numerical results have demonstrated the stability and robustness in generating the formations.

Reading the Media’s Mind

Abstract: ‘There’s no art to find the mind’s construction in the face,’ wrote Shakespeare, but trying to infer what someone is really thinking is arguably the essence of interaction between cognitive agents. Turning this into a computational model is challenging, but one possible approach to infer mental models from linguistic expression is to look at patterns of lexical associations. Assuming that written language reflects conceptual associations in the writer’s mind, we have previously shown differences in the patterns of lexical association between creative and non-creative writing. In this paper, we apply the same approach to news reports from individual media sources over the same period, with the goal of looking for differential associative patterns. The underlying assumption is that the associative patterns of a media source will reflect its “mind” and “personality,” i.e., specific styles, preferences, or biases, just as they do for individuals.

Mining the Temporal Structure of Thought from Text

Abstract: Thinking is a self-organized dynamical process and, as such, interesting to characterize. However, direct, real-time access to thought at the semantic level is still very limited. The best that can be done is to look at spoken or written expression. The question we address in this research is the following: Is there a characteristic pitch of thought? To begin answering this complex question, we look at text documents from several large corpora at the sentence level – i.e., using sentences as the units of meaning – and considering each document to be the result of a random process in semantic space. Given a large corpus of multi-sentence documents, we build a lexical association network representing associations between words in the corpus. This network is used to induce a semantic similarity metric between sentences, and each document is segmented into multi-sentence semantically coherent blocks (SCBs) with occasional connecting text between the blocks. Based on this segmentation, the process of document generation is modeled as a sticky Markov chain at the sentence level. We show that most documents across all the corpora are sequences of blocks with a very consistent mean length of 6.4 sentences across the corpora. This consistency suggests that a value of 6-7 sentences may be the typical mean length for single coherent thoughts in texts. We have also described several ways of visualizing the semantic structure of documents in space and time.