I am a doctoral candidate in RAIR Lab in Computer Science department of Rensselaer Polytechnic Institute. My advisor is Prof. Selmer Bringsjord. and I am funded by ONR MURI. My dissertation is on Detection of Counter-masquerading using Counteridenticals. I did my Bachelors in Indian Institute of Science and Technology Shibpur in Computer Science.
Research
My areas of interest are Logic, Natural Language Processing and generation, Ethics in AI, Emotions and Reasoning, fraud and security in social networks. Most of my work involves using logical framework and modern Natural language processing tools to explore ethical aspects, both in classical AI and social networks.
Projects and Publications
| Project Name | Description | Publications, Posters, Talks |
|---|---|---|
| Counteridenticals and counter-masquerading | Counteridenticals are a special type of Counterfactuals that deal with identity. It is the key to formalizing the question "What if X were Y?" The formalization of counteridenticals is much different from counterfactuals, and we further use this formalization directly to counter-masquerading. We have applied our theories to social media where people create fake accounts to impersonate real or imaginary people. | Poster: Novel algorithms of Counter-masquerading using Counterfactuals, Ghosh R., Bringsjord S., ICRES 2018 |
| Publication: Deontic Counteridenticals Bringsjord, S., Ghosh, R. & Payne-Joyce, J. (2016) “Deontic Counteridenticals and the Design of Ethically Correct Intelligent Agents: First Steps,” in Bonnet, G., Haarbers, M., Hindriks, K., Katell, M. & Tessier, K. Proceedings of the 1st Workshop on Ethics in the Design of Intelligent Agents (Proc. EDIA 2016), pp. 38–43. Workshop held in The Hague, Holland, August 30 2016. |
||
| Presentation : Extracting Creatures of Fiction for Story-Based QA (S-BQA), Bringsjord, S., Ghosh, R., Govindarajulu, N.S., Licato, J., August 2 2019, Boston, MA, at Story Enabled Intelligence, a workshop at the 2019 edition of Advances in Cognitive Systems | ||
| Emotion and Reasoning | We introduce a new emotion model Felmë, rooted in DCEC and based on OCC. This is a highly expressive affective model which can be applied easily with state of the art Natural language systems. We have applied this model in ethical robotics and counter-masquerading. | Publication : Toward the Engineering of Virtuous Machines Govindarajulu, N.S., Bringsjord, S., Ghosh, R., & Sarathy, V. (2019) “Toward the Engineering of Virtuous Machines.” AAAI/ACM Conference on Artificial Intelligence, Ethics, and Society (AIES 2019). |
| Presentation: Extremely Emotional Robots Presented at the July 17–21 2017 Focused Session Philosophical Perspectives in the Technology, Consciousness conference series sponsored and co ̈ordinated by SRI, in Menlo Park, CA. Bringsjord presented, on July 19 2017 |
||
| Poster : A Theory of Emotions in a Cognitive Calculus, Ghosh R., Govindarajulu N.S., Scally S., Bringsjord S., Advances in Cognitive Systems conference 2017. | ||
| Ethical robotics and natural language | For scenarios where robots are supposed to work with humans or complement human activities, the presence of an ethical system is highly necessary. The works in this section encompasses application of affective reasoning as well as natural language applications in the Ethical robotics. | Publication : Beyond the Doctrine of Double Effect : A Formal Model of True Self-Sacrifice Govindarajulu, N.S., Bringsjord, S., Ghosh, R. & Peveler (2019) “Beyond the Doctrine of Double Effect: A Formal Model of True Self-Sacrifice” in Ferreira, M.I.A., Sequeira, J.S., Virk, G.S., Tokhi, M.O., Kadar, E.E., eds., Robots and Well-Being, in the series Intelligent Systems, Control and Automation: Science and Engineering (Basel, Switzerland: Springer), pp. 39–54. |
| Upcoming Publication : Bringsjord, S., Licato, J., Ghosh, R., Bello, P., Bridewell, W., Payne-Joyce, J. “The Interrogation Room” Minds and Machines. | ||
| Miscellaneous Natural language applications | Natural Language applications based on logic and machine learning has been used in various projects for a seamless communication with the robots or simulated artificial agents. | Publication : Toward a smart city using Tentacular AI Sen, A., Bringsjord, S., Govindarajulu, N.S., Mayol, P., Ghosh, R., Srivastava, B. & Talamadupula, K. (2018) “Toward a Smart City Using Tentacular AI” in Kameas, A. & Stathis, K., eds., Proceedings of the 14th European Conference on Ambient Intelligence (AMI) (Basel, Switzerland: Springer Nature AG), pp. 106–112. This volume is in Lecture Notes in Computer Science, Vol. 11249. |
| Publication : Real Robots that pass the test of Self Consciousness Bringsjord, S., Licato, J., Govindarajulu, N.S., Ghosh, R. & Sen, A. (2015) “Real Robots that Pass Tests of Self-Consciousness” in Proceedings of the 24th IEEE Inter- national Symposium on Robot and Human Interactive Communication (RO-MAN 2015), (New York, NY: IEEE), pp. 498–504. |
||
| Poster : Peveler M., O'Neill K., Sen A., Ghosh R., Dong R., Bringsjord S.,, "The Planning Dilemma in Cognitive Computing for CISL's Immersive 'Cognitive Boardroom'", Cognitive Colloquium on Augmenting Human Intelligence, IBM 2016 |
Demonstration Videos
Extremely emotional Robots
This demonstration shows an extremely emotional robot which gets angered by external stimuli and led by emotions to commit something which is not a desirable action. Thankfully, the robot realizes that and becomes remorseful of its actions. This whole interaction is based on Felmë Theory of emotions which portray how emotions are generated in a robot.
Nao bots and Suicide
This demonstration shows different views of Nat bots regarding suicide. We have formalized two different philosophical views regarding suicide, and showed how using counteridenticals, the Nao comes to its own decision regarding the issue.
AI in an interrogation room
This simple demonstration shows how AI is able to derive the implicit knowledge of the members in an interrogation room, and help the interrogator with the new knowledge.
Converting Natural Language commands into actions
Given a natural language command to perform a series of actions, this parser captures the semantics of the actions and converts it into a series of procedures required to successfully accomplish the goal of the actions intended in a given simulation environment.
Relevance Parsing
In addition to the tools we have been developing for parsing of statements in natural language (NL), we have been addressing problems related to deeper semantic parsing. The problem of relevance is well known to experts in many fields: computational linguistics, analogy, and legal reasoning, just to name a few. If a commander were to say the following to a robot:
“If you need the med kit, it will be at location bravo.”
We would want the robot to commit the knowledge “the med kit is at location bravo” to memory. However, at first glance, the original statement was phrased in the form of a material conditional, meaning that a simple approach might instead represent this knowledge in an if-then form that was not originally intended. This is a problem, because the robot’s knowledge of the med kit’s location should not be contingent on whether or not he needs it!