Citation Context

...gative) of each opinion sentence. This consists of two steps: (1) for each opinion word in the opinion word list, we identify its semantic orientation using a bootstrapping technique and the WordNet (=-=Miller et al. 1990-=-), and (2) we then decide the opinion orientation of each sentence based on the dominant orientation of the opinion words in the sentence. The details are presented in a subsequent paper. Experiments ...

Citation Context

...atistical approaches that exploiting the fact that the words composing a term tend to be found close to each other and reoccurring (Jacquemin and Bourigault 2001; Justeson and Katz 1995; Daille 1996; =-=Church and Hanks 1990-=-). However, using noun phrases tends to produce too many non-terms, while using reoccurring phrases misses many low frequency terms, terms with variations, and terms with only one word. Our associatio...

Citation Context

...of terms, namely noun phrases, and statistical approaches that exploiting the fact that the words composing a term tend to be found close to each other and reoccurring (Jacquemin and Bourigault 2001; =-=Justeson and Katz 1995-=-; Daille 1996; Church and Hanks 1990). However, using noun phrases tends to produce too many non-terms, while using reoccurring phrases misses many low frequency terms, terms with variations, and term...

Citation Context

...dentification. The majority of text summarization techniques fall in two categories: template instantiation and text extraction. Work in the former framework includes (DeJong 1982), (Tait 1983), and (=-=Radev and McKeown 1998-=-). They focus on the identification and extraction of certain core entities and facts in a document, which are packaged in a template. This framework requires background analysis to instantiate a temp...

Citation Context

...e document. Over the years, many sophisticated techniques were developed, e.g., strong notions of topicality (Hovy and Lin 1997), lexical chains (Barzilay and Elhadad 1997), and discourse structures (=-=Marcu 1997-=-). Our work is different as we do not extract those most representative sentences, but only identify and extract those specific product features and the opinions related with them. Kan and McKeown (19...

Citation Context

...rases, and statistical approaches that exploiting the fact that the words composing a term tend to be found close to each other and reoccurring (Jacquemin and Bourigault 2001; Justeson and Katz 1995; =-=Daille 1996-=-; Church and Hanks 1990). However, using noun phrases tends to produce too many non-terms, while using reoccurring phrases misses many low frequency terms, terms with variations, and terms with only o...

Citation Context

... summarization and terminology identification. The majority of text summarization techniques fall in two categories: template instantiation and text extraction. Work in the former framework includes (=-=DeJong 1982-=-), (Tait 1983), and (Radev and McKeown 1998). They focus on the identification and extraction of certain core entities and facts in a document, which are packaged in a template. This framework require...

Citation Context

...nces of “autofocus” are replaced with “auto-focus”. Frequent Features Generation This step is to find features that people are most interested in. In order to do this, we use association rule mining (=-=Agrawal and Srikant 1994-=-) to find all frequent itemsets. In our context, an itemset is a set of words or a phrase that occurs together. Association rule mining is stated as follows: Let I = {i1, …, in} be a set of items, and...

Citation Context

... sentence. The reason is that other components of a sentence are unlikely to be product features. Here, pre-processing includes the deletion of stopwords, stemming and fuzzy matching. Fuzzy matching (=-=Jokinen and Ukkonen 1991-=-) is used to deal with word variants or misspellings. For example, “autofocus” and “auto-focus” actually refer to the same feature. All the occurrences of “autofocus” are replaced with “auto-focus”. F...

Citation Context

...identify and extract those specific product features and the opinions related with them. Kan and McKeown (1999) propose a hybrid approach that merges template instantiation with sentence extraction. (=-=Boguraev and Kennedy 1997-=-) also reports a technique that finds a few very prominent expressions, objects or events in a document and use them to help summarize the document. Again, our work is different as we need to find all...

Citation Context

...ties and facts in a document, which are packaged in a template. This framework requires background analysis to instantiate a template to a suitable level of detail. It is thus not domain independent (=-=Sparck-Jones 1993-=-a, 1993b). Our technique does not fill in any template and is domain independent. The text extraction framework (Paice 1990; Kupiec, Pedersen, and Chen 1995; Hovy and Lin 1997) identifies some represe...

Citation Context

...entifies some representative sentences to summarize the document. Over the years, many sophisticated techniques were developed, e.g., strong notions of topicality (Hovy and Lin 1997), lexical chains (=-=Barzilay and Elhadad 1997-=-), and discourse structures (Marcu 1997). Our work is different as we do not extract those most representative sentences, but only identify and extract those specific product features and the opinions...

Citation Context

... following the <individual reviews> link to see why existing customers like it or what they complain about. Our task is clearly different from traditional text summarization (Radev and McKeown. 1998; =-=Hovy and Lin 1997-=-) in a number of ways. First of all, our summary is structured rather than another (but shorter) free text document as produced by most text summarization systems. Second, we are only interested in fe...

Citation Context

...ties and facts in a document, which are packaged in a template. This framework requires background analysis to instantiate a template to a suitable level of detail. It is thus not domain independent (=-=Sparck-Jones 1993-=-a, 1993b). Our technique does not fill in any template and is domain independent. The text extraction framework (Paice 1990; Kupiec, Pedersen, and Chen 1995; Hovy and Lin 1997) identifies some represe...