Viewing 5 reply threads
  • Author
    Posts
    • #47633
      Anastasios Asimakopoulos
      Keymaster
          @anastasios

          Here, you can ask any technical questions you might have about the function we focussed on, Frequency. We can clarify things from the guides and/or the practical tasks e.g. if you got different results, couldn’t perform the search, need help to understand your word sketch, etc. We can also answer any questions that arose during any other searches you might have performed on your own for practice.

        • #51824
          Paula Acejo Cantero
          Participant
              @powla

              Hello,

               

              I found very interesting the exploration by fields / domains. Expression more widely used in biology due to gene expression than in linguistics, and attack more in Medicine anc IT than in History…. it contradicted some of my expectations.

            • #51827
              Paula Acejo Cantero
              Participant
                  @powla

                  I used ngram and found an obvious trend in BAWE we find adverbial expressions or impersonal verbal structures whereas in BASE the most frequent 4grams rrlate to personal expressions such as I dont you, you would like and similar. This reflects reality, as academic language is supposed to be more scientific and impersonal.

                  • #51847
                    Anastasios Asimakopoulos
                    Keymaster
                        @anastasios

                        Thank you @powla for sharing your observations with us. It’s great to see you confirming what you already know about academic language as well as discovering new surprising findings. Your comment about the style of academic language made think wonder how scientific/impersonal academic speaking is. To find out, I created two n-gram lists, one using BASE (academic speaking) and one using the BNC 2014 Spoken (informal speaking). Here, I am attaching the top 20 4-grams with some similarities and differences I identified:

                        • The lexical chunks I don’t know, I don’t think and don’t want to are found in both top 20 lists. Interestingly, I don’t know is almost 5 times more frequent and I don’t think is almost 4 times more frequent in BNC than in BASE. (yellow highlights)
                        • Many 4-grams in the BNC top 20 include the pronoun I (purple highlights), which is not the case for the BASE top 20. Could it be that speakers in informal conversations refer to themselves more often than speakers in academic settings? It might be worth looking into the frequency of the pronoun in the two corpora in general.
                        • BASE top 20 includes 4 n-grams with the pronoun you, while BNC includes 2
                        • BASE top 20 also includes some n-grams that we are likely to find in academic writing (red highlights), while BNC 2014 Spoken does not include these in the top 20. It would be worth checking the relative frequencies of these in BAWE and BNC2014 Spoken too.

                         

                        • Comparison of top 20 4-grams BASE vs BNC2014 Spoken

                        I hope you found this mini-exploration useful ;)

                    • #52631
                      Prisila Mlingi
                      Participant
                          @prisila

                          I do not have questions but I was able to attempt most of the tasks  and get the correct answers ;-)

                          Would I be correct to say that, we can use this function of sketch engine to teach academic lexical bundles?

                          • #52636
                            Anastasios Asimakopoulos
                            Keymaster
                                @anastasios

                                Thank you for your question @prisila Yes that’s correct. The Frequency function can help us summarise patterns from our concordance lines, so yes we can examine sequences of any size to the left, right or both sides of our search term. By the way, I saw your post in the Self-discovery forum, so it might help to know that n-grams can be used to examine lexical bundles too ;)

                            • #52867
                              Siti Asmiyah
                              Participant
                                  @siti-asmiyah

                                  I selected the task of ‘your own research study’ as I am interested on finding’lexicogrammatical pattern.’ I did a research on lexicogrammatical pattern in the past and use other application to generate the pattern from the corpora I collected. With this Sketch Engine, I was going fine with the search of the patterns. I find it interesting when searching for the word list in  METCLIL and BASE. Although the two take academic discourse of spoken English and both take transcripts from seminar, the frequency of the words used, they have slightly different top 10 adjectives on the list. What I really like from this Sketch Engine is that we can download the data in xls format that enables me to compare across different group of corpora.

                                • #52872
                                  Anastasios Asimakopoulos
                                  Keymaster
                                      @anastasios

                                      Thank you for sharing your search with us @siti-asmiyah If you are interested in comparing the frequencies of words in these two lists, you can click on the View options button (the eye button) and turn on the toggle Frequency per million. This will make the frequencies of words comparable since they are normalised using the same basis (per million). For example, we can see that the 1st person singular pronoun i (I) occurs 26,917.63 times per million in the EMI Seminars corpus, but 12,320.78 times per million in the BASE corpus.

                                  Viewing 5 reply threads
                                  • You must be logged in to reply to this topic.