Feedback system is an essential part of any academic institute for maintaining and monitoring the academic quality . In an academic institution, students feedback is one of the most powerful influences on learning. A student can give their feedback about the courses they have taken and the faculties whom they are taught. Student feedback can play a vital role in ensuring the quality of the education .
Williams, Cappuccini, Ansfield  have collected feedback from students about their experiences in institutions has become one of the central pillars of the quality process. The collection and publication of students’ feedback now provide a key element in many processes of quality assurance and enhancement. This was indeed a welcome development because students now are expected to play a more active role in the learning processes.
Brennan and Williams  discussed that there are two principal reasons for collecting feedback from students. The first is to enhance the students’ experience of teaching and learning and the second is to contribute to the monitoring and review of quality and standards. They later identified seven other reasons behind the collection of student data, which range from ensuring the effectiveness of course design and delivery to contributing to staff or faculty development.
Harvey  observed that the most important use of student feedback is in providing senior management with invaluable information from the student’s perspective to assist in an institution’s continuous quality improvement process. Thus, a tertiary institution’s assessment process would be incomplete without having this aspect of feedback. The author believed that there are six main reasons why feedback is needed. These include the fact that feedback can: provide information for improvement, provide information for prospective students, provide information for current students, address accountability issues, provide benchmarking information and used to make comparisons between and within institutions.
Traditional student course evaluation feedback systems are pen and paper-based where it generates a huge amount of data and hence makes the feedback analysis very difficult . Student feedback (SF) can be positive or negative toward a course or a faculty, so sentiment analysis(SA) can be applied for classification of the SF. Sentiment analysis (SA) purposes to identify the attitude of a speaker or a writer with respect to some topics. SA has been applied in many areas such as product reviews , movie reviews , restaurant reviews , and Twitter . Pang and Lee  expanded the basic task of categorizing a movie review as either positive or negative to predicting star ratings on either a 3 or 4-star scale, while Snyder and Barzilay  executed a details analysis of restaurant reviews, calculating ratings for different aspects such as the food and environment (on a five-star scale) of a specified restaurant.
Sentiment analysis (also known as opinion extraction) refers to utilize of natural language processing, text analysis and computational linguistics to find and extract particular information from texts. The output of sentiment analysis is identifying the polarity of texts as positive, negative or neutral toward a certain topic. In SFC system, texts collected from open-ended questions can be valuable feedback to an institutional organization. Sentiment analysis techniques can generally have classified into machine learning approach and lexicon-based approach. Existing famous sentiment analysis methods include machine learning, lexicon and hybrid approach.
The machine learning approach (ML) applies supervised or unsupervised learning with collective of linguistic features for prediction. The supervised methods make utilize of a big number of labeled training data. The unsupervised methods are utilized when it is hard to find these labeled training data. ML approach is usually applied for a big dataset such as for reviews of different contexts . While effective, ML approach needs big dataset which is labor demanding to collect.
Lexicon based approach is a base on a sentiment lexicon, a group of known and precompiled sentiment terms. It can be classified into dictionary-based approach and corpus-based approach. The dictionary-based approach initially searches opinion words, and then searches the dictionary of their synonyms and antonyms. The corpus-based approach begins with a group of opinion words, and then finds other opinion words in a big corpus to help in finding opinion words in specific context orientations. Dictionary-based approach is finding opinion word which is time-consuming work which can slow the processing of the system and it also has constraint which is the inability to find opinion words with external domain and context specific orientations . Therefore, there is a possibility of missing opinion words in organizational culture dimension if only utilized dictionary-based approach. Ahmed and Hoda  stated that utilizing the corpus based-approach only is not as effective as the dictionary-based approach because it is difficult to prepare a big corpus to cover all English words. Though it is performed well in context-specific orientation like Twitter .
A hybrid approach is more effective than utilizing a single approach for opinion analysis. Generally, the hybrid approach had targeted to build based on specific context orientation and it’s not effective in outside of its own orientation  .
Qiu et al.  utilized a lexicon-based dictionary-based approach for measuring sentiment sentences in contextual advertising. They recommended an advertising approach for better ads relevance and user experience. In the research, Chetan and Atul  utilized a lexicon-based approach such that a dictionary of sentiment contains opinion words were used to categorize the text into positive, negative or neutral statement. Authors noted machine learning approaches are not effective for their case because it takes much time doing sentiment analysis as they have to be trained first. According to Chetan and Atul, their method performed very well in terms of accuracy. Tweets were analyzed with an accuracy of 73.5 % and time taken 14.8 seconds for total 6,74,412 tweets.
In hybrid technique, both combinations of machine learning and lexicon based approaches are utilized and the technique can provide improve performance according to researchers . Though, Walaa, Ahmed and Hoda  stated that the hybrid approach’s computational difficulty is complex compared to use a single method of SA. Therefore, a hybrid approach is not easy to build. Mudinas, Zhang and Levene  have shown that improved performance of classification in a hybrid approach than only using machine and lexicon-approaches. They proposed a concept-level sentiment analysis system, called pSenti, which is built by combining lexicon and machine learning approaches which benefit of world stability and high accuracy. Authors  introduced a sentence level, sentiment polarity calculation that classifies complex sentential structures and adjusts the sentence polarity accordingly. They used English dictionary and set of heuristic rules for all the sentences in a review, gives the overall polarity of the review. They found the result is more accurate in many cases and in other cases the results are the same. The combination of a rule-based classifier and a supervised learning with lexicon English dictionary has proposed for tweet analysis . The tweets are classified as positive, negative and unknown.
Most of the previous works of students’ feedback systems are positive or negative feedback findings. In this paper, we propose a hybrid technique which is based on sentiment analysis and our system can give not only the polarity of the student feedback but also can give the suggestion for a course or a faculty.
This essay has been submitted by a student. This is not an example of the work written by our professional essay writers. You can order our professional work here.