Milind Pande

Bio: Milind Pande is an academic researcher from Massachusetts Institute of Technology. The author has contributed to research in topics: Wireless sensor network & Efficient energy use. The author has an hindex of 4, co-authored 8 publications receiving 44 citations.

Intelligent Twitter Spam Detection: A Hybrid Approach

Abstract: Over the years there has been a large upheaval in the social networking arena. Twitter being one of the most widely-used social networks in the world has always been a key target for intruders. Privacy concerns, stealing of important information and leakage of key credentials to spammers has been on the rise. In this paper, we have developed an Intelligent Twitter Spam Detection System which gives the precise details about spam profiles by identifying and detecting twitter spam. The system is a Hybrid approach as opposed to single-tier, single-classifier approaches which takes into account some unique feature sets before analyzing the tweets and also checks the links with Google Safe Browsing API for added security. This in turn leads to better tweet classification and improved as well as intelligent twitter spam detection.

H2E2: A hybrid, hexagonal & energy efficient WSN green platform for precision agriculture

Abstract: To enhance productivity of farms there is need to follow mantra by Food and Agricultural Organization of the United Nations (FAO) i.e. “Save & Grow”. Save and Grow directs towards Precision Agriculture (PA) for sustainable crop production intensification. There is challenge to establish information communication technology (ICT) technologies in fields and remote areas. Proposed system is novel platform to establish wireless sensor network (WSN) to execute PA overcoming the challenge to set up wireless sensor and energy efficient network at farms. Hybrid network of free space optics (FSO) and radio frequency (RF) connectivity and use of solar at farms and base station / control station makes it energy efficient. License free, high bandwidth FSO technology enables multimedia data exchange thus opening gate for many applications. Hybrid Hexagonal Positioning (HHP) for sensor nodes has achieved better link utilization compared to other topologies saving energy and increasing lifetime of sensor node and network. Overall energy efficient, centralized, environment friendly system allows cost effective solution to farmers. A village centric model is presented to define applicability of proposed solution.

Energy efficient hybrid architecture and positioning of sensors in WSN for precision agriculture

Abstract: Establishing Wireless Sensor Networks (WSN) for precision agriculture has become need of time now but few challenges like energy, reliability, cost etc. need to overcome and serve the application in most efficient manner. Considering advantage of open space i.e. Line of Sight (LOS) in agriculture fields and solar presence this paper proposes the novel architecture of WSN which uses existing Radio Frequency/Free Space Optical (RF/FSO) link model and hybrid energy source for base station for precision agriculture. Randomly deployed sensors consume more energy than organized sensor deployment hence positioning of sensor is critical factor for energy efficiency of WSN, this paper also proposes sensor positioning method for agriculture fields which covers maximum area and minimum energy. Use of available best routing protocol in the application will add more energy efficiency to the system. Reduction in number of sensor nodes to maximum area will also optimize the cost of project. Overall creating win-win situation to establish WSN for precision agriculture is the objective of this paper.

The role of intraocular lens exchange in the management of major implant-related complications.

Abstract: This study evaluates the role of intraocular lens exchange procedures in the management of major implant-related complications. Medical records of 30 patients undergoing intraocular lens exchange at the Leeds General Infirmary from 1 January 1984 to 30 November 1991 were reviewed. Details of their primary implantation surgery, intervening ophthalmic history, lens exchange surgery and outcome were analysed. Follow-up ranged from 3 to 210 weeks. Seventy-six per cent of cases achieved final visual acuity of 6/12 or better. Visual acuity improved in 60%, worsened in 6.6% and was unchanged in 33.4% compared with pre-operative levels. Three patients had cystoid macular oedema, 3 ocular hypertension, 2 bullous keratopathy, 1 chronic anterior uveitis and 1 patient developed a retinal detachment. This approach to managing major implant-related complications gives good visual results. There is a significant complication rate but the risk-benefit ratio justifies the use of the technique.

Study of Patents Filed in India in Mechanical Engineering Sector

Abstract: In the present study, an attempt has been made to identify the trends of patenting activity in mechanical engineering in India using data from value added databases. It is a first time full data compilation of patent filing and granted data at Indian Patent Office from year 1961 and patent filing data in the different areas of inventions from 1986. Further for this study a separate search was conducted using patent information databases with focus (IPC based) on mechanical engineering sector to understand and identify the companies working in this area. More than 10,000 patent applications and grants exclusively related to mechanical engineering have been considered for analysis. The study indicates focus of research of different organizations/industries in specific areas of mechanical engineering, R&D collaboration between the companies. The paper also addresses various trends and discusses the R&D partnership observed in Indian organizations and foreign organizations. The study further highlights the important directions for future path of R&D.

Evolving Support Vector Machines using Whale Optimization Algorithm for spam profiles detection on online social networks in different lingual contexts

Abstract: A new classification approach based Support Vector Machine is proposed for detecting spammers on Twitter.The proposed approach reveals the most influencing features in the process of identifying spammers.Different lingual contexts are studied: Arabic, English, Spanish, and Korean. Detecting spam profiles is considered as one of the most challenging issues in online social networks. The reason is that these profiles are not just a source for unwanted or bad advertisements, but could be a serious threat; as they could initiate malicious activities against other users. Realizing this threat, there is an incremental need for accurate and efficient spam detection models for online social networks. In this paper, a hybrid machine learning model based on Support Vector Machines and one of the recent metaheuristic algorithms called Whale Optimization Algorithm is proposed for the task of identifying spammers in online social networks. The proposed model performs automatic detection of spammers and gives an insight on the most influencing features during the detection process. Moreover, the model is applied and tested on different lingual datasets, where four datasets are collected from Twitter in four languages: Arabic, English, Spanish, and Korean. The experiments and results show that the proposed model outperforms many other algorithms in terms of accuracy, and provides very challenging results in terms of precision, recall, f-measure and AUC. While it also helps in identifying the most influencing features in the detection process.

Wireless Sensor Network and Internet of Things in Precision Agriculture

Abstract: Internet of Things is one of the most popular subjects nowadays where sensors and smart devices facilitate the provision of information and communication. In IoT, one of the main concepts is wireless sensor networks in which data is collected from all the sensors in a network characterized by low power consumption and a wide range of communication. In this study, an architecture to monitor soil moisture, temperature and humidity on small farms is provided. The main motivation for this study is to decrease water consumption whilst increasing productivity on small agricultural farms and precisions on them. This motivation is further propelled by the fact that agriculture is the backbone of some towns and most villages in most of the countries. Furthermore, some countries depend on farming as the main source of income. Putting the above-mentioned factors into consideration, the farm is divided into regions; the proposed system monitors soil moisture, humidity and temperature in the respective regions using wireless sensor networks, internet of things and sends a report to the end user. The report contains, as part of the information, a 10-day weather forecast. We believe that with the above information, the end user (farmer) can more efficiently schedule farm cultivation, harvesting, irrigation, and fertilization.

Spam review detection using spiral cuckoo search clustering method

Abstract: Nowadays, online reviews play an important role in customer’s decision. Starting from buying a shirt from an e-commerce site to dining in a restaurant, online reviews has become a basis of selection. However, peoples are always in a hustle and bustle since they don’t have time to pay attention to the intrinsic details of products and services, thus the dependency on online reviews have been hiked. Due to reliance on online reviews, some people and organizations pompously generate spam reviews in order to promote or demote the reputation of a person/product/organization. Thus, it is impossible to identify whether a review is a spam or a ham by the naked eye and it is also impractical to classify all the reviews manually. Therefore, a spiral cuckoo search based clustering method has been introduced to discover spam reviews. The proposed method uses the strength of cuckoo search and Fermat spiral to resolve the convergence issue of cuckoo search method. The efficiency of the proposed method has been tested on four spam datasets and one Twitter spammer dataset. To validate the efficacy of proposed clustering method it is compared with six metaheuristics clustering methods namely; particle swarm optimization, differential evolution, genetic algorithm, cuckoo search, K-means, and improved cuckoo search. The experimental results and statistical analysis validate that the proposed method outruns the existing methods.