Deck 7: Text Analytics, Text Mining, and Sentiment Analysis

Question 1

In the financial services firm case study, text analysis for associate-customer interactions were completely automated and could detect whether they met the company's standards.

Answer 1

True

Question 2

Current use of sentiment analysis in voice of the customer applications allows companies to change their products or services in real time in response to customer sentiment.

Answer 2

True

Question 3

In text mining, inputs to the process include unstructured data such as Word documents, PDF files, text excerpts, e-mail and XML files.

Answer 3

True

Question 4

Regional accents present challenges for natural language processing.

Question 5

Categorization and clustering of documents during text mining differ only in the preselection of categories.

Question 6

In sentiment analysis, sentiment suggests a transient, temporary opinion reflective of one's feelings.

Question 7

In the patent analysis case study, text mining of thousands of patents held by the firm and its competitors helped improve competitive intelligence, but was of little use in identifying complementary products.

Question 8

The bag-of-words model is appropriate for spam detection but not for text analytics.

Question 9

In the chapter's opening vignette, IBM's computer named Watson outperformed human game champions on the game show Jeopardy!

Question 10

In the Hong Kong government case study, reporting time was the main benefit of using SAS Business Analytics to generate reports.

Question 11

In text mining, if an association between two concepts has 7% support, it means that 7% of the documents had both concepts represented in the same document.

Question 12

In sentiment analysis, it is hard to classify some subjects such as news as good or bad, but easier to classify others, e.g., movie reviews, in the same way.

Question 13

In the BBVA case study, text analytics was used to help the company defend and enhance its reputation in social media.

Question 14

The linguistic approach to speech handles processes elements such as intensity, pitch and jitter from speech recorded on audio.

Question 15

In text mining, creating the term-document matrix includes all the terms that are included in all documents, making for huge matrices only manageable on computers.

Question 16

Text analytics is the subset of text mining that handles information retrieval and extraction, plus data mining.

Question 17

Articles and auxiliary verbs are assigned little value in text mining and are usually filtered out.

Question 18

Chinese, Japanese, and Thai have features that make them more difficult candidates for natural language processing.

Question 19

Detecting lies from text transcripts of conversations is a future goal of text mining as current systems achieve only 50% accuracy of detection.

Question 20

During information extraction, entity recognition (the recognition of names of people and organizations) takes place after relationship extraction.

Question 21

What types of documents are BEST suited to semantic labeling and aggregation to determine sentiment orientation?

A) medium- to large-sized documents
B) small- to medium-sized documents
C) large-sized documents
D) collections of documents

Question 22

What do voice of the market (VOM) applications of sentiment analysis do?

A) They examine customer sentiment at the aggregate level.
B) They examine employee sentiment in the organization.
C) They examine the stock market for trends.
D) They examine the "market of ideas" in politics.

Question 23

All of the following are challenges associated with natural language processing EXCEPT

A) dividing up a text into individual words in English.
B) understanding the context in which something is said.
C) distinguishing between words that have more than one meaning.
D) recognizing typographical or grammatical errors in texts.

Question 24

In text analysis, what is a lexicon?

A) a catalog of words, their synonyms, and their meanings
B) a catalog of customers, their words, and phrase
C) a catalog of letters, words, phrases and sentences
D) a catalog of customers, products, words, and phrase

Question 25

How is objectivity handled in sentiment analysis?

A) It is ignored because it does not appear in customer sentiment.
B) It is incorporated as a type of sentiment.
C) It is clarified with the customer who expressed it.
D) It is identified and removed as facts are not sentiment.

Question 26

According to a study by Merrill Lynch and Gartner, what percentage of all corporate data is captured and stored in some sort of unstructured form?

A) 15%
B) 75%
C) 25%
D) 85%

Question 27

In text mining, stemming is the process of

A) categorizing a block of text in a sentence.
B) reducing multiple words to their base or root.
C) transforming the term-by-document matrix to a manageable size.
D) creating new branches or stems of recorded paragraphs.

Question 28

In sentiment analysis, which of the following is an implicit opinion?

A) The hotel we stayed in was terrible.
B) The customer service I got for my TV was laughable.
C) The cruise we went on last summer was a disaster.
D) Our new mayor is great for the city.

Question 29

In the Whirlpool case study, the company sought to better understand information coming from which source?

A) customer transaction data
B) delivery information
C) customer e-mails
D) goods moving through the internal supply chain

Question 30

In text mining, tokenizing is the process of

A) categorizing a block of text in a sentence.
B) reducing multiple words to their base or root.
C) transforming the term-by-document matrix to a manageable size.
D) creating new branches or stems of recorded paragraphs.

Question 31

Identifying the target of an expressed sentiment is difficult for all the following reasons EXCEPT

A) the review may not be directly connected to the target through the topic name.
B) blogs and articles with the sentiment may be general in nature.
C) strong sentiments may be generated by a computer, not a person.
D) sometimes there are multiple targets expressed in a sentiment.

Question 32

In the opening vignette, the architectural system that supported Watson used all the following elements EXCEPT

A) massive parallelism to enable simultaneous consideration of multiple hypotheses.
B) an underlying confidence subsystem that ranks and integrates answers.
C) a core engine that could operate seamlessly in another domain without changes.
D) integration of shallow and deep knowledge.

Question 33

Which of these applications will derive the LEAST benefit from text mining?

A) patients' medical files
B) patent description files
C) sales transaction files
D) customer comment files

Question 34

Inputs to speech analytics include all of the following EXCEPT

A) written transcripts of calls to service centers.
B) recorded conversations of customer call-ins.
C) live customer interactions with service representatives.
D) videos of customer focus groups.

Question 35

What application is MOST dependent on text analysis of transcribed sales call center notes and voice conversations with customers?

A) finance
B) OLAP
C) CRM
D) ERP

Question 36

In the research literature case study, the researchers analyzing academic papers extracted information from which source?

A) the paper abstract
B) the paper keywords
C) the main body of the paper
D) the paper references

Question 37

Sentiment classification usually covers all the following issues EXCEPT

A) classes of sentiment (e.g., positive versus negative).
B) range of polarity (e.g., star ratings for hotels and for restaurants).
C) range in strength of opinion.
D) biometric identification of the consumer expressing the sentiment.

Question 38

In text mining, which of the following methods is NOT used to reduce the size of a sparse matrix?

A) using a domain expert
B) normalizing word frequencies
C) using singular value decomposition
D) eliminating rarely occurring terms

Question 39

What data discovery process, whereby objects are categorized into predetermined groups, is used in text mining?

A) clustering
B) association
C) classification
D) trend analysis

Question 40

In the Blue Cross Blue Shield case study, speech analytics were used to identify "confusion" calls by customers. What was true about these calls?

A) They took less time than others as frustrated customers hung up.
B) They led customers to rely more on self-serve options.
C) They were not documented by customer service reps for speech analytics.
D) They were difficult to identify using standard phrases like "I don't get it."

Question 41

________ is probably the most often used form of information extraction.

Question 42

At a very high level, the text mining process can be broken down into three consecutive tasks, the first of which is to establish the ________.

Question 43

In the text mining system developed by Ghani et al., treating products as sets of ________ rather than as atomic entities can potentially boost the effectiveness of many business applications.

Question 44

When identifying the polarity of text, the most granular level for polarity identification is at the ________ level.

Question 45

In automated sentiment analysis, two primary methods have been deployed to predict sentiment within audio: acoustic/phonetic and ________ modeling.

Question 46

________ is mostly driven by sentiment analysis and is a key element of customer experience management initiatives, where the goal is to create an intimate relationship with the customer.

Question 47

When viewed as a binary feature, ________ classification is the binary classification task of labeling an opinionated document as expressing either an overall positive or an overall negative opinion.

Question 48

When a word has more than one meaning, selecting the meaning that makes the most sense can only be accomplished by taking into account the context within which the word is used. This concept is known as ________.

Question 49

In sensitivity analysis, the task of differentiating between a fact and an opinion can also be characterized as calculation of ________ polarity.

Question 50

When labeling each term in the WordNet lexical database, the group of cognitive synonyms (or synset) to which this term belongs is classified using a set of ________, each of which is capable of deciding whether the synset is Positive, or Negative, or Objective.

Question 51

Where ________ appears in text, it comes in two flavors: explicit, where the subjective sentence directly expresses an opinion, and implicit, where the text implies an opinion.

Question 52

________ focuses on listening to social media where anyone can post opinions that can damage or boost your reputation.

Question 53

________ is a technique used to detect favorable and unfavorable opinions toward specific products and services using large numbers of textual data sources.

Question 54

Because the term-document matrix is often very large and rather sparse, an important optimization step is to reduce the ________ of the matrix.

Question 55

In the Mining for Lies case study, a text based deception-detection method used by Fuller and others in 2008 was based on a process known as ________, which relies on elements of data and text mining techniques.

Question 56

IBM's Watson utilizes a massively parallel, text mining-focused, probabilistic evidence-based computational architecture called ________.

Question 57

Among the significant advantages associated with the ________ approach to linguistic modeling is the method's ability to maintain a high degree of accuracy no matter what the quality of the audio source, and its incorporation of conversational context through the use of structured queries.

Question 58

The time-demanding and laborious process of the ________ approach makes it impractical for use with live audio streams.

Question 59

________ models operate on the premise that, when in a charged state, a speaker has a higher probability of using specific words, exclamations, or phrases in a particular order.

Question 60

________, also called homonyms, are syntactically identical words with different meanings.

Question 61

Natural language processing (NLP), a subfield of artificial intelligence and computational linguistics, is an important component of text mining. What is the definition of NLP?

Question 62

When IBM Research began looking for a major research challenge to rival the scientific and popular interest of Deep Blue, the computer chess-playing champion, what was the company's goal?

Question 63

Identify, with a brief description, each of the four steps in the sentiment analysis process.

Question 64

Name and briefly describe four of the most popular commercial software tools used for text mining.

Question 65

Sentiment analysis has many names. Which other names is it often known by?

Question 66

In the security domain, one of the largest and most prominent text mining applications is the highly classified ECHELON surveillance system. What is ECHELON assumed to be capable of doing?

Question 67

Describe the query-specific clustering method as it relates to clustering.

Question 68

What is the definition of text analytics according to the experts in the field?

Question 69

How would you describe information extraction in text mining?

Question 70

Within the context of speech analytics, what does the linguistic approach focus on?