Conference Articles:


2017:

[1] Relatedness-based Multi-Entity Summarization, Kalpa Gunaratna, Amir Hossein Yazdavar ,Krishnaprasad Thirunarayan ,Amit Sheth, and Gong Cheng (Accepted IJCAI 2017).

Representing world knowledge in a machine processable format is important as entities and their descriptions have fueled tremendous growth in knowledge-rich information processing platforms, services, and systems. In this research, we present an approach that can summarize facts about a collection of entities by analyzing their relatedness in preference to summarizing each entity in isolation.


[2] Semi-Supervised Approach to Monitoring Clinical Depressive Symptoms in Social Media, Amir Hossein Yazdavar, Hussein S. Al-Olimat, Monireh Ebrahimi, Tanvi Banerjee, Krishnaprasad Thirunarayan, Jyotishman Pathak and Amit Sheth, (Accepted ASONAM 2017).

Using huge set of tweets crawled from users with self-declared depression symptoms, our study uses a novel, semi-supervised LDA-based model to monitor depression symptoms through their expression on Twitter (in terms of word usage patterns and topical preferences) to emulate PHQ-9 depression recognition system. Our screening tool is able to identify clinical depressive symptoms with an accuracy of 68% and precision of 72%.

[3] Identifying Depressive Disorder in the U.S. Population using Twitter Data (In prepreation)

Clinical depression is a serious challenge in public health worldwide. While successful early identification and treatment can contribute to many positive health and behavioral benefits, depression remains untreated or undertreated due to several reasons including the denial of illness or the social stigma. With the prolific use of social media platforms, millions of people express and share their moods, feelings and daily struggles with mental health issues routinely. Insights gleaned from social media such as Twitter can be regarded as complementary to the current surveybased methods that can assist both governmental and nongovernmental organizations in policy development. In this study, we examined exploitation of big (social) data for identification of depression and its population trend in different regions. This paper presents the curation of one of the largest datasets of depressed individuals’ profiles studied to date. In particular, using statistical techniques along with heterogeneous sets of features, we developed regionspecific models to automatically detect depressed individuals along with their geographical information on Twitter and link them to US Census data through their location. Our findings from the geographical analysis of social media correlate with depression statistics reported by the Substance Abuse and Mental Health Services Administration (SAMHSA). We also evaluate the performance of our system in terms of the average F-score by testing it against human judgment.

[4] Mining Adverse Drug Reaction in Social Media via Deep Recurrent Neural Networks and Background Knowledge, Amir Hossein Yazdavar, Hussein S. Al-Olimat, Monireh Ebrahimi, Goonmeet Bajaj, Krishnaprasad Thirunarayan, Jyotishman Pathak and Amit Sheth, (In preparation for TACL 2017).

Adverse drug reactions (ADRs) are considered major public health challenge and the sixth leading cause of death worldwide. Unwanted drug effects also have significant economic aswell as clinical costs as they often lead to hospital admissions. Consequently, we introduce DeepADRMine, a semantic annotator tool which extract medical concepts from unstructured text by integrating lexicon-based method (top-down processing) and data-driven method (bottom-up processing). In particular, leveraging available ADR lexicon (FAERS) we first extract explicit and non-descriptive ADRs. Then, to further enhance the semantic understand- ing of user generated contents, we treat the implicit ADRs (descriptive) detection as a task of sequence labeling.

Journal Articles:

2017:

[1] On the Difficulty of Sentiment Analysis for Dynamic Events: Presidential Election, Monireh Ebrahimi, Amir Hossein Yazdavar, Amit Sheth, IEEE Inteligent Systems, Impact Factor:2.34, Accepted

[2] Identifying Pragmatic Functions in Social Media Indicative of Gender-Based Violence Beliefs, Tanvi Banerjee, Amir Hossein Yazdavar, Hemant Purohit, Andrew Hampton, Valerie L. Shalin, Amit P. Sheth, (In Preparation to submit to Computer in Human Behavior)

2016:

[1] Fuzzy Based Implicit Sentiment Analysis on Quantitative Sentences, Amir Hossein Yazdavar, Monireh Ebrahimi, Naomie Salim, Journal of Soft Computing and Decision Support Systems.

[2] Recognition of Side Effects as Implicit-Opinion Words in Drug Reviews, Monireh Ebrahimi, Amir Hossein Yazdavar, Naomie Salim, Safaa Eltyeb, Online Information Review, Impact Factor: 1.15.

[3] Rock strength estimation: a PSO-based BP approach, E.Tonnizam Mohamad, D. Jahed Armaghani, E. Momeni, Amir Hossein Yazdavar, Monireh Ebrahimi, Journal of Neural Computing and Applications, Impact Factor: 1.49.

2014:

1] Transmission of Data with OFDM Technique for Communication Networks Using GHz Frequency Band Soliton Carrier, Iraj Sadegh Amiri, Monireh Ebrahimi, Amir Hossein Yazdavar, S. Ghorbani, S. E. Alavi, Sevia M. Idrus , J. Ali, IET Communications Journal (IEEE), Impact Factor: 0.74.

[2] Analytical Modeling and Simulation of IV Characteristics in Carbon Nanotube based Gas Sensors using ANN and SVR Methods, Elnaz Akbari, Zolka- fle Buntat, Aria Enze- vaee,Monireh Ebrahimi, Amir Hossein Yazdavar, Rubiyah Yusof, Chemometrics and Intelligent Laboratory Systems, Impact Factor: 2.32.