Ashish Saini

A new fuzzy logic based ranking function for efficient Information Retrieval system

Abstract The relevant documents from large data sets are retrieved with the help of ranking function in Information Retrieval system. In this paper, a new fuzzy logic based ranking function is proposed and implemented to enhance the performance of Information Retrieval system. The proposed ranking function is based on the computation of different terms of term-weighting schema such as term frequency, inverse document frequency and normalization. Fuzzy logic is used at two levels to compute relevance score of a document with respect to the query in present work. All the experiments are performed on CACM and CISI benchmark data sets. The experimental results reveal that the performance of our proposed ranking function is much better than the fuzzy based ranking function developed by Rubens along with other widely used ranking function Okapi-BM25 in terms of precision, recall and F-measure.

Fuzzy logic-based approach to develop hybrid similarity measure for efficient information retrieval

A similarity measure is used in information retrieval systems to retrieve and rank the relevant documents. In this paper, a new fuzzy-based approach to develop hybrid similarity measure is proposed and implemented. The proposed approach overcomes the limitations of extensively used similarity measures such as Cosine, Jaccard, Euclidean and Okapi-BM25 along with Genetic Algorithm-based hybrid similarity measures proposed by researchers. This approach uses fuzzy rules to infer the weights of different similarity measures. In this paper, the experiments are performed on CACM and CISI benchmark data collections. The performance of the proposed approach is evaluated in terms of precision, recall and average precision and average recall of retrieved relevant documents. The results are compared with different similarity measures available in literature. The results show the marked improvement in performance of information retrieval systems using the proposed fuzzy logic-based hybrid similarity measure.

A Novel Hybrid Fuzzy Multi-Objective Evolutionary Algorithm: HFMOEA

This paper presents a development of a new hybrid fuzzy multi-objective evolutionary algorithm (HFMOEA) for solving complex multi-objective optimization problems. In this proposed algorithm, two significant parameters such as crossover probability (P C) and mutation probability (P M) are dynamically varied during optimization based on the output of a fuzzy controller for improving its convergence performance by guiding the direction of stochastic search to reach near the true pareto-optimal solution effectively. The performance of HFMOEA is examined and compared with NSGA-II on three benchmark test problems such as ZDT1, ZDT2 and ZDT3.

A novel Fuzzy-PSO term weighting automatic query expansion approach using combined semantic filtering

Abstract Information Retrieval system retrieves relevant documents from large datasets. Automatic Query Expansion (AQE) is one of the approaches to enhance IR performance by adding additional terms to original query. The selection of suitable additional terms for AQE is a crucial task. Term weighting method is one of the ways to deal with such a problem. This paper presents a new term weighting based AQE approach to retrieve more relevant documents from data corpus. The proposed approach comprises of three major steps. First step determines the optimal weights of different IR evidences for different terms using Particle Swarm Optimization (PSO). Fuzzy logic technique is used to improve performance of PSO by controlling inertia and acceleration coefficients during the optimization. Co-occurrence score is introduced as new IR evidence in the proposed approach. Second step is focused on removal of noisy terms by using new combined semantic filtering method. Third step reweights the terms using Rocchio method. The proposed approach is compared with recently developed automatic query expansion approaches in terms of performance measures such as precision, recall, F-measure and MAP (Mean Average Precision). Three benchmark datasets CACM, CISI and TREC - 3 are used to verify the results. The proposed approach is found better than other approaches according to results obtained for these benchmark datasets.

Fuzzy Logic Based Similarity Measure for Information Retrieval System Performance Improvement

The documents of any information retrieval system are ranked on the basis of similarity measure. Some similarity measures e.g. Cosine, Euclidean and Okapi etc. have been extensively used for retrieving relevant documents against the query. In present paper, a new fuzzy based similarity measure is proposed. Experiments have been performed on CACM data collection. The performance of proposed similarity measure is evaluated and compared with above mentioned similarity measures on the basis of Precision-Recall curves, average similarity value of documents for individual query and average number of retrieved relevant documents. The results show the marked improvement in performance of information retrieval system using proposed fuzzy logic based similarity as compare to other similarity measures.

Multi-objective reactive power market clearing in competitive electricity market using HFMOEA

This paper presents an application of a hybrid fuzzy multi-objective evolutionary algorithm (HFMOEA) for solving a highly constraint, mixed integer type, complex multi-objective reactive power market clearing (RPMC) problem for the competitive electricity market environment. In HFMOEA based multi-objective optimization approach, based on the output of a fuzzy logic controller crossover and mutation probabilities are varied dynamically. It enhances stochastic search capabilities of HFMOEA. In multi-objective RPMC optimization framework, two objective functions namely the total payment function (TPF) for reactive power support from generators and synchronous condensers and the total real transmission loss (TRTL) are minimized simultaneously for clearing the reactive power market. The proposed HFMOEA based multi-objective RPMC scheme is tested on a standard IEEE 24 bus reliability test system and its performance is compared with five other multi-objective evolutionary techniques such as MOPBIL, NSGA-II, UPS-EMOA and SPEA-2 and a new extended form of NSGA (ENSGA-II). Applying all these six evolutionary techniques, a detailed statistical analysis using T-test and boxplots is carried out on three performance metrics (spacing, spread and hypervolume) data for RPMC problem. The obtained simulation results confirm the overall superiority of HFMOEA to generate better Pareto-optimal solutions with higher convergence rate as compared to above mentioned algorithms. Further, TPF and TRTL values corresponding to the best compromise solutions are obtained using said multi-objective evolutionary techniques. These values are compared with one another to take better market clearing decisions in competitive electricity environment.

Fuzzy Based Approach to Develop Hybrid Ranking Function for Efficient Information Retrieval

Ranking function is used to compute the relevance score of all the documents in document collection against the query in Information Retrieval system. A new fuzzy based approach is proposed and implemented to construct hybrid ranking functions called FHSM1 and FHSM2 in present paper. The performance of proposed approach is evaluated and compared with other widely used ranking functions such as Cosine, Jaccard and Okapi-BM25. The proposed approach performs better than above ranking functions in terms of precision, recall, average precision and average recall. All the experiments are performed on CACM and CISI benchmark data collections.

A new similarity function for information retrieval based on fuzzy logic

In this paper, a novel approach is presented to construct a similarity function to make information retrieval efficient. This approach is based on different terms of term-weighting schema like term frequency, inverse document frequency and normalization. The proposed similarity function uses fuzzy logic to determine similarity score of a document against a query. All the experiments are done with CACM benchmark data collection. The experimental results reveal that the performance of proposed similarity function is much better than the fuzzy based ranking function developed by Rubens along with other widely used similarity function Okapi-BM25 in terms of precision rate and recall rate.

A new swarm-based efficient data clustering approach using KHM and fuzzy logic

AbstractClustering is a useful technique to create different groups of objects on the basis of their nature. Objects of same group are of similar in nature and differ to the objects of other groups. Clustering has proved its importance in various fields such as information retrieval, bioinformatics, image processing and many others. In this paper, particle swarm optimization (PSO) technique is used with K-harmonic means (KHM) for clustering. PSO overcomes the limitations of KHM like local optimum problem. Fuzzy logic is also employed in this paper to make PSO adaptive in nature by controlling various parameters. The performance of the proposed approach is validated on five benchmark datasets in terms of inter-clustering distance, intra-clustering distance, F-measure and fitness value. The results of proposed approach are compared with well-known conventional clustering techniques such as K-means, KHM and fuzzy C-means along with different state-of-the-art clustering approaches. Two text-based benchmark datasets such as CACM and CISI are also used to test the performance of all clustering approaches. The proposed clustering approach gives better results in comparison with other clustering approaches as clear from both the experimental and statistical analyses.

Multi-objective congestion management based on generator's real & reactive power rescheduling bids in competitive electricity markets

In this paper, a competitive bidding based congestion management strategy is developed for a pool type electricity market model. The proposed congestion management strategy alleviates the transmission congestion by using rescheduling both the real and reactive power output of generators. The separate bids for real and reactive powers are invited from all participating generators in the congestion market. Further, the multi-objective optimization frameworks are developed which minimizes two contradictory objective functions such as Total Cost of Congestion Management and Congestion Severity Index (CSI) subjected to various system constraints. The proposed strategy is tested for different congestion scenarios on 39-bus New England power system.