Rule Mining Techniques Research Articles

Exploring Students’ Feedback in Online Assessment System Using Opinion Mining Technique

Opinion mining has been widely used in recent online reviews or feedback due to its ability to analyse text-based data. The use of this technique for analysing data from students’ feedback needs to be addressed, since most educational institutions are focusing more on questionnaires based on the Likert-scale rather than on the open-review type. To this end, there is a lack of online assessment systems that could automatically analyse open-review questionnaires. Therefore, the main aim of this study is to analyse students’ feedback in an online assessment system through the opinion mining technique, by focusing on textual form data derived from the open-review questionnaires. To achieve this aim, an opinion mining feedback system, known as the OMFeedback, was developed. The Vader Sentiment Intensity Analyser was adapted for processing students’ feedback and the lexicon based approach was used for analysing the words. In addition, the OMFeedback incorporates the capitalisation of words and emoji features to enrich the capability of the system. This newly developed system could lead to new paradigms in educational institutions for enhancing students’ learning process and for guiding them through their learning journey.

Rule-Based Effective Collaborative Recommendation Using Unfavorable Preference

The biggest challenge in collaborative filtering recommendation system research is data sparsity ; it mainly occurs as user rates very few items from widely available options. Data Imputation techniques address the data sparsity problem by filling the missing values and then predicting the likeliness of the user. Most of the existing imputation systems assign high ratings to the items or incorporate additional information to enhance the performance of collaborative filtering recommendations. This paper proposes an association rule-based imputation method (RUBLE) to improve the top- $N$ prediction performance of the collaborative filtering recommendation. The proposed method identifies the unfavorable items of each user using the association rule mining technique and imputes them with low values. The proposed method not only addresses the sparsity problem but also provides a better quality of recommendation by eliminating the unfavorable items in top- $N$ predictions. Existing collaborative methods can quickly adapt to the proposed method as it is method agnostic. The experimental results show that the proposed method enhances the accuracy of the traditional recommender methods by two times on average and significantly outperforms existing imputation based approaches.

Classification of Pruning Methodologies for Model Development using Data Mining Techniques

Knowledge discovery process deals with two essential data mining techniques, association and classification. Classification produces a set of large number of associative classification rules for a given observation. Pruning removes unnecessary class association rules without losing classification accuracy. These processes are very significant but at the same time very challenging. The experimental results and limitations of existing class association rules mining techniques have shown that there is a requirement to consider more pruning parameters so that the size of classifier can be further optimized. Here through this paper we are presenting a survey various strategies for class association rule pruning and study their effects that enables us to extract efficient compact and high confidence class association rule set and we have also proposed a pruning methodology

An assessment of opinions and perceptions of smart thermostats using aspect-based sentiment analysis of online reviews

Smart thermostats have been on the market for nearly a decade, with an estimated adoption rate of 7% in 2018. With many regions of the U.S. having a heating and/or cooling system in nearly 100% of households, there is significant opportunity for further adoption, which can help support energy savings and building-grid interactions. However, more insight is needed to provide a better understanding of their utilization and user opinions. In this study, online reviews are used to evaluate users’ perceptions and attitudes towards smart thermostats. 26,372 product reviews were collected for five commercially-available smart thermostats and were analyzed with a confirmatory aspect-based opinion mining technique. An analysis of this dataset shows that the characteristics of the current user population show substantial differences compared to the more widely studied early adopters. When comparing the most commonly discussed topics, users generally do not discuss the energy and cost savings related features of their devices in comparison to other topics such as control, ease of use, and installation. In addition, comfort is discussed nearly twice as much as energy efficiency. The results of this work can help product manufacturers and utility providers to push towards more widespread adoption and efficient use.

Opinion mining technique for developing student feedback analysis system using lexicon-based approach (OMFeedback)

Online assessment systems are increasingly utilised as an evaluation tool for measuring the performance of lecturers in Institutions of Higher Learning (IHLs). These systems commonly have a set of questionnaires comprised of quantitative and qualitative questions. Most online lecturer teaching assessment systems are focused on the quantitative part of the questionnaire since it is easy to analyse. On the contrary, the qualitative part, which requires students’ opinions, is often omitted and neglected, and the level of opinion results are excluded. This is because students’ opinions are usually in the form of unstructured texts, which makes it hard to manually analyse all the feedback. To address this problem, a system that could automatically analyse students’ feedback (known as OMFeedback) was developed. This system applies an opinion mining technique to reveal the teaching assessment results, which are underpinned by a lexicon-based approach. Lexicon-based is a common textual data quantification method that can analyse the sentiment tendency of a student’s feedback. A database of English sentiment words, known as the Vader Lexicon, was utilised as a lexical source to determine the polarity of words. By analysing these sentiment words, which included intensifier words extracted from students’ feedback, this system was able to determine the results for the lecturer teaching assessment by describing the level of positive, negative, or neutral opinions. This system was also able to process new features, such as capitalised words and emoji characters to enhance the opinion results. Ultimately, this newly developed system will provide useful information to IHLs for improving lecturers’ teaching skills and course materials.

Automated discovery of declarative process models with correlated data conditions

Automated process discovery techniques enable users to generate business process models from event logs extracted from enterprise information systems. Traditional techniques in this field generate procedural process models (e.g., in the BPMN notation). When dealing with highly variable processes, the resulting procedural models are often too complex to be practically usable. An alternative approach is to discover declarative process models, which represent the behavior of the process as a set of constraints. Declarative process discovery techniques have been shown to produce simpler models than procedural ones, particularly for processes with high variability. However, the bulk of approaches for automated discovery of declarative process models focus on the control-flow perspective, ignoring the data perspective. This paper addresses the problem of discovering declarative process models with data conditions. Specifically, the paper tackles the problem of discovering constraints that involve two activities of the process such that each of these two activities is associated with a condition that must hold when the activity occurs. The paper presents and compares two approaches to the problem of discovering such conditions. The first approach uses clustering techniques in conjunction with a rule mining technique, while the second approach relies on redescription mining techniques. The two approaches (and their variants) are empirically compared using a combination of synthetic and real-life event logs. The experimental results show that the former approach outperforms the latter when it comes to re-discovering constraints artificially injected in a log. Also, the former approach is in most of the cases more computationally efficient. On the other hand, redescription mining discovers rules with higher confidence (and lower support) suggesting that it may be used to discover constraints that hold for smaller subsets of cases of a process.

Implementation of Client Controlled Privacy Preserving Data Model for Mining Decision Rules Using Decision Tree and Association Rules

The privacy-preserving data mining (PPDM) is one of the techniques which are used for mining data dynamically with preserving privacy of the end data owner. In this paper, a PPDM technique for generating the privacy-preserving decision rules is proposed and implemented. The key motive of presenting this privacy-preserving decision rule mining technique is to demonstrate how securely data is aggregated in the PPDM environment, how securely extract them and consumed them with the help of applications. In addition to comparing the state of art methods for mining privacy preserving decision rules for preparing the future directions of research. Therefore two different data models have used namely decision tree and association rule mining. The conducted experiments demonstrate that decision tree-based techniques are superior to the association rule mining based techniques for mining higher dimensional data with higher accuracy and low resource consumption. Therefore in the near future for extending this data model the two concepts are also introduced in this paper.

A method for extracting performance metrics using work-order data

Holistic performance metrics are necessary to understand how operational resources are used and to detect anomalous zones, floors, equipment, and work-order categories in large commercial and institutional buildings. Work-order data in computerized maintenance management systems (CMMS) represent an untapped potential to extract such performance metrics. In this paper, a method to conduct text analytics on CMMS data is developed and demonstrated through a case study in which four years’ worth of data from four large commercial buildings are used. Association rule mining technique is employed to identify building, system, and component-level recurring work-order taxonomies and common failure modes. The results highlight the potential of kernel density functions, decision trees, Sankey diagrams, survival curves and stacked line plots to effectively visualize the temporal, spatial, and categorical anomalies in the complaint patterns. It is identified that often only a few floors and complaint types account for most of the complaints in a building. The analysis of operator comments reveal that the most frequent lighting-related complaints are resolved by replacing ballasts and lights, and the thermal and air quality complaints are addressed by adjusting the temperature setpoints, airflow rates, and fan operation schedules.

Examining differences among book reviews from various online platforms

Purpose Websites have their own features in aspect preference (e.g. the relative importance platforms place on product aspects in product evaluation). The purpose of this paper is to capture characteristics of different book reviews on aspect preferences by opinion mining techniques. Design/methodology/approach The authors employ two indicators for identifying aspect preferences, and propose a method for quantifying overall differences of reviews on aspect preferences through three dimensions: aspect awareness, aspect satisfaction and comprehensive value. Findings The results show that book reviews on e-commerce websites contain information about external aspects of a book (e.g. hardcover), while those on social network websites pay more attention to content-related aspects of the book (e.g. stories). These results indicate that aspect preferences of reviews vary from platforms and make it hard to evaluate book comprehensively based on single-source data. Online book reviews from a wide range of sources can assess book impact from multiple perspectives and dimensions. Practical implications In order to illustrate the value of the authors’ method, the authors show book impact assessment based on multi-source data as an application of these difference analyses. Furthermore, the authors present an example of a book promotion to provide customized marketing services for different user clusters. Originality/value This study investigates the influence of different data sources on book evaluation from the content of book reviews. The authors also showcase potential applications of these analyses in book impact assessment.

EARMGA and Apriori Algorithm's Performance Evaluation for Association Rule Mining

Association rule mining techniques are important part of data mining to derive relationship between attributes of large databases. Association related rule mining have evolved huge interest among researchers as many challenging problems can be solved using them. Numerous algorithms have been discovered for deriving association rules effectively. It has been evaluated that not all algorithms can give similar results in all scenarios, so decoding these merits becomes important. In this paper two association rule mining algorithms were analyzed, one is popular Apriori algorithm and the other is EARMGA (Evolutionary Association Rules Mining with Genetic Algorithm). Comparison of these two algorithms were experimentally performed based on different datasets and different parameters like Number of rules generated, Average support, Average Confidence, Covered records were detailed.

Sentiment Research on Student Feedback to Improve Experiences in Blended Learning Environments

Educational data mining has aroused a great research interest among the educational institutions nowadays. Blended learning is used as a pedagogy in the field of teaching and learning. Blended learning is a fusion of online digital media with conventional teaching in classrooms where the teacher and student has to be present physically throughout teaching duration. Teacher –student interaction is made possible using the internet during the non-contact hours. For rendering a valuable blended learning environment, it is essential to gather users’ opinion or feedback on this learning methodology. Therefore, opinion-mining techniques have been used in this paper for helping the academicians to improve and promote such learning environments. Students positive or negative feelings towards the subject teaching can be analyzed using these techniques. This paper discusses how sentiment analysis can be performed on the feedback collected in a learning management system with the intention of advancing the teaching learning process. This work presents the experimental results that were obtained after comparison of various feature selection methods say Chi-square, Information Gain, Mutual Information and Symmetrical Uncertainty

Analysis of Policy Changes and User Satisfaction of Road Transportation Services using Opinion Mining Techniques

Analysis of Policy Changes and User Satisfaction of Road Transportation Services using Opinion Mining Techniques

A data‐mining‐based approach to informed decision‐making in engineering education

AbstractIn outcome‐based academic programs, Program Educational Objectives (PEOs) and Student Outcomes (SOs) are two cores around which all programs’ components and processes revolve. Needless to say, the PEOs‐SOs mapping is very critical for program success and, therefore, a deep understanding of PEOs, SOs, and their intra/inter‐correlations is very important for effective program's decisions making. In this context, this paper proposes a data mining‐based approach to discover hidden knowledge of PEOs, SOs, and their mapping and correlations in engineering programs. More specifically, the proposed approach employs association rule mining techniques to generate association rules, among and between PEOs and SOs, from PEOs‐SOs mapping data of a set of engineering programs, which can be then filtered and manipulated to discover the desired knowledge. To this end, a set of 152 self‐study reports of engineering programs, accredited by American Board for Engineering and Technology‐Engineering Accreditation Commission (ABET‐EAC), are collected and the mapping data between their PEOs and ABET‐EAC SOs are extracted. The dataset is processed and transformed into a representation suitable for association rules mining. This involves developing a set of PEOs labels, annotating data with PEOs labels, and extracting three target datasets. Apriori algorithm is then applied to each dataset to generate three sets of association rules. The generated association rules are then filtered and manipulated to discover the knowledge of PEOs, ABET‐EAC SOs, and their mapping and correlations. Finally, a discussion on the informativeness of the discovered knowledge for the decisions making in engineering education is given.

Adverse event profiles of ifosfamide-induced encephalopathy analyzed using the Food and Drug Administration Adverse Event Reporting System and the Japanese Adverse Drug Event Report databases.

Ifosfamide is extensively used to treat several malignant conditions. Administration of ifosfamide can cause encephalopathy and other neurotoxic effects. The aim of this study was to obtain novel information on the onset profiles of ifosfamide-induced encephalopathy (IIE) considering other associated clinical factors using the US Food and Drug Administration Adverse Event Reporting System (FAERS) and the Japanese Adverse Drug Event Report (JADER) databases. We analyzed the reports of encephalopathy between 2004 and 2018 from the FAERS and JADER databases. To define IIE, we used the Medical Dictionary for Regulatory Activities (MedDRA) preferred terms and standardized queries. The reporting odds ratios (ROR) at 95% confidence interval (CI) was used to detect the signal for IIE and adjusted for covariates using a multivariate logistic regression technique. We evaluated the time-to-onset profile of IIE and used the association rule mining technique to discover undetected associations, such as potential risk factors. In the FAERS database, the ROR (CI) for encephalopathy (preferred term, PT) and encephalopathy (standardized MedDRA queries, SMQ) was 56.58 (51.69-61.93) and 1.57 (1.48-1.67), respectively. In the JADER database, the ROR (95% CI) for encephalopathy (PT) and encephalopathy (SMQ) was 13.54 (9.91-18.50) and 1.24 (1.01-1.53), respectively. The multivariate logistic regression analysis showed a significant contribution in IIE signal in the ≥ 60year group (p = 0.00094; vs. < 60year group) and ≥ 2000mg/m2 dosage group (p = 0.00045; vs. < 2000mg/m2 dosage group). The association rules of {ifosfamide, aprepitant} → {encephalopathy (SMQ)} demonstrated high lift values. The average dose of ifosfamide in patients with encephalopathy (PT) and without encephalopathy (PT) was 2022.8 ± 592.8 (mean ± standard deviation) and 1568.5 ± 703.2mg/m2, respectively (p < 0.05). Encephalopathy within the first 7days of ifosfamide administration was 94.1% for encephalopathy (PT) and 87.7% for encephalopathy (SMQ), respectively. The present analysis demonstrated that the incidence of encephalopathy with ifosfamide should be closely monitored for a short onset (within 7days). The patients who are administered a high dose of ifosfamide or co-administrated aprepitant should be carefully monitored for the development of encephalopathy.

Cluster Centric Technique for Association Rule Mining(C-ARM) For High Precision and Coverage in Web Page Recommendation

Association Rule Mining (ARM) is known for its popularity and efficiency in the data mining domain. Over the recent years, the amount of data that gets accumulated in the internet is getting increased exponentially over time. The data available so are stored in online and are retrieved when a user requests for the same through key words with the help of a search engine. The important task of the search engines are to present the appropriate web pages that an user is expecting and in the modern times, The need of the hour is to recommend web pages to the users that he is interested in. This made the web page recommendation an important and vital task. Although many of the researchers are in the preliminary task of developing such systems, we in this research propose a recommendation model in which different users are interested upon a common item or domain by using the ARM concept. The data patterns that are in common are identified using the ARM and further these are clustered on a form of hierarchy .The clusters makes the recommendation system to easily identify the user group and based on the group, the pages are recommended, The experimental analysis are discussed and found to be efficient than the available methods in terms of computation time and reliability.

Sistem Rekomendasi Buku untuk Perpustakaan Perguruan Tinggi Berbasis Association Rule

One of the services in the university library is an information system to find the availability of library collections and the location of each collection shelf. But not many of these systems provide a mechanism that can recommend visitors not only about the books they want, but also other related books that may be needed. This study uses association rule mining techniques that are applied to library transaction data to identify relationships between books (titles) that attract visitors' attention. Relationships are built on interesting measurements between the titles, namely support and confidence, where support determines the combination of the most frequently borrowed book titles, while confidence produces the possibility that the title of the book will be borrowed along with other books. The pattern of book titles association with high confidence indicates that the titles are very related so it is recommended for visitors to consider borrowing along with the book they are looking for. In addition, the system can also recommend the procurement of new books and rack configurations to improve the visitor's experience when searching for books on the site. In the experiment, the precision of recommendations generated from the system reached 70%. Web applications were developed to help understand the effectiveness of the recommendation system based on association rules.

Combined Approach for Teachers’ Evaluation Aspects Identification Using Dictionary and Patterns Based

Teacher performance evaluation is a common method and often used for evaluates teaching quality in higher education. With the rapid growth of opinion mining technique. Aspect-based opinion mining application has been possibly employed to extraction and summarization of students' comments for teacher evaluation. However, to automated teacher evaluation features identification from a large number of students' comments collection is very hard work. This study has the goal to address this problem. The main objectives of the proposed method are: (1) to identify teacher evaluation aspects, (2) to compare the efficiency of dictionary based, patterns based and the combination of them, and (3) to enhance the accuracy result in the teachers’ evaluation aspects identification from the unstructured text of students' feedbacks. The students' feedbacks were collected by questionnaires and the dataset was constructed manually with a total of 4,496 sentences from 300 undergraduate student responses in 10 subjects by purposive sampling and the collection of positive and negative sentences from 30 participants group interviewed in the workshop. Both approaches were applied to identify the frequency teachers' evaluation aspects. The experimental results found that our proposed approach provided reasonably more accurate results, the combination approach enhanced a significantly average of precision and recall. For future work, we focus on the application of new linguistic patterns and non-frequency aspects in order to increase the accuracy result.&#x0D; Keywords—aspects identification, lexicon relation, linguistic pattern, opinion mining, teacher evaluation.

Identifying the Underlying Relationship Between Water Quality Parameters of the Groundwater Samples using Association and Clustering Algorithms in Coimbatore District

Water is a highly complex environmental system; its protection cannot be met by traditional methods. As a part of the process, it is mandatory to evaluate the parameters of ground water so as to pursue suitable treatment. These days’ data mining algorithms have been developed to handle various data-rich environmental problems. In data mining, several techniques such as complex non-linear science, soft computing techniques, clustering and association have been applied in the domain of ground water quality assessment and evaluation in and around Coimbatore District. In this work, the statistical cluster analysis methods and association rule mining techniques were used to identify the spatial distribution of different cluster of wells having similar characteristics and determine the relationship between different water quality variables. The water quality assessment in Coimbatore was done using 13 parameters, namely NO3 - , TDS, Mg2+, Ca2+, Na+ , Cl- , F- , SO4 2- , EC, pH and Hardness including location in different sites. The main objective of the present study is to assess the performance of various clustering algorithms of WEKA and identify the most suitable algorithm for clustering water quality samples. K-Mean algorithm and centroid method of Hierarchical clustering performed in the similar manner in clustering. In addition to that, this study focused on identifying the water quality parameters exceeding permissible limits that occur together (TDS, Mg2+, SO4 2- , EC, hardness) in the given samples using Association Algorithms. The performance and efficiency of different association algorithms like Apriori and Frequent Pattern Growth algorithm was evaluated by factors like support, confidence, lift, leverage and conviction values.

A hybrid recommender system for the mining of consumer preferences from their reviews

Product review sites are widespread on the Internet and are rapidly gaining in popularity among consumers. This already large volume of user-generated content is dramatically growing every day, making it hard for consumers to filter out the worthwhile information which appears on the various review sites. There commendation system plays a significant role in solving the problem of information overload. This study proposes a framework which integrates a collaborative filtering approach and an opinion mining technique for movie recommendation. Within the proposed framework, sentiment analysis is first applied to the users’ reviews to detect consumer opinions about the movie they have watched and to explore the individual’s preference profile. Traditional recommendation models are overly dependent on preference ratings and often suffer from the problem of ‘data sparsity’. Experimental results obtained from real online reviews show that our proposed method is effective in dealing with insufficient data and is more accurate and efficient than existing traditional methods.

Development of an Intelligent Job Recommender System for Freelancers using Client’s Feedback Classification and Association Rule Mining Techniques

Development of an Intelligent Job Recommender System for Freelancers using Client’s Feedback Classification and Association Rule Mining Techniques